Chapter 6: Co-Designing a Remote Communication System. ...... the strategies and tools families use to support remote communication, and the role remote ...
REFLECTIVE DESIGN-IN-USE: CO-DESIGNING AN ASSISTIVE REMOTE COMMUNICATION SYSTEM WITH INDIVIDUALS WITH COGNITIVE DISABILITIES AND THEIR FAMILIES
by MELISSA DAWE B.A., University of California at Berkeley, 1998 M.S., University of Colorado, 2001
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirement for the degree of Doctor of Philosophy Department of Computer Science 2007
This thesis entitled: Reflective Design-In-Use: Co-Designing an Assistive Remote Communication System with Individuals with Cognitive Disabilities and their Families written by Melissa Dawe has been approved for the Department of Computer Science
_______________________________ (Gerhard Fischer)
_______________________________ (Tamara Sumner)
Clayton Lewis
Leysia Palen
Katie Siek
James Sullivan
Kathleen Tierney
Date__________________ The final copy of this thesis has been examined by the signatories, and we find that both the content and the form meet acceptable presentation standards of scholarly work in the above mentioned discipline. HRC protocol # 1202.14, # 0206.43
Dawe, Melissa (Ph.D., Computer Science) Reflective Design-In-Use: Co-Designing an Assistive Remote Communication System with Individuals with Cognitive Disabilities and their Families Thesis directed by Professor Gerhard Fischer
Mobile phones have transformed the way we communicate with friends and family, coordinate our daily activities, and organize our lives. For families with children with cognitive disabilities there is widespread hope, though not always fulfilled, that personal technologies – particularly mobile phones – can bring a dramatic increase in their child’s level of safety, independence, and social connectedness. However, many individuals with significant cognitive disabilities are unable to use off-the-shelf mobile phones because of a mismatch between their needs and abilities and the technology design at the user interface and functionality level. In this research, I conducted ethnographic studies to first understand the current patterns of use and the adoption process of assistive technology for individuals with cognitive disabilities and their families. I identified barriers to technology adoption including the lack of consideration given to a caregiver network, lack of simplicity, and lack of support in upgrade and replacement of technology. I then combined participatory design and a “technology probe” study with individual families with a young adult with cognitive disabilities to co-design a PDAbased mobile phone that matched the abilities and needs of their child. This study included an eight-week field trial with the phone to evaluate the user interface and functionality. The technology evolved during the field study based on the experiences of use. This research identified requirements for functionality and customization of mobile phones for this population. The importance of emotional and social motivations emerged as an important focus in the design of assistive technology. This method also proved to be an effective way to engage individuals with cognitive disabilities as participants in the design of their own technology.
iii
To My Family My source of strength, compassion, humor and love
Acknowledgements I am very fortunate that the long journey getting my doctorate was never traveled alone. First and foremost, I would not be here if not for my mother and father, Amy Birdsong Dawe and John Dawe. I thank them for their steadfast support, love, and confidence that I can achieve whatever I set my heart on; and for instilling me with a deep sense of curiosity and intellectual inquiry. I thank my three sisters Michri, Valerie, and Amanda –amazing women and life-long allies who, through our shared world view, make me feel sane (and if I am crazy, at least I’m in great company). I am lucky to have an amazingly loving and fun extended family on the Birdsong, Dawe, and McCormick sides. In particular I thank George and Anne Birdsong for encouraging me to seek and pursue success in what I love. I would like to thank the members of my dissertation committee, especially my advisor Gerhard Fischer, for guiding me along the way while encouraging me to develop and pursue my own ideas. I thank the entire group at L3D for providing an intellectually stimulating and supportive environment, and for challenging me to always push myself and explore new avenues in my research. I thank the many inspiring women in computer science I have met along the way, especially the crew at Georgia Tech who supported and included me as one of their own at many conferences. I thank Bill and Claudia Coleman for generously funding this research through the Coleman Foundation. I thank Dr. David Braddock at the Coleman Institute for supporting and guiding my work. The open, trusting and long-term nature of their financial support enabled my research to focus on what my research participants said was important, rather than on a hypothesis conceived a priori in a research lab. I have also received financial support from the Beverly Sears Graduate Student Award, and travel grants from the Institute of Cognitive Science and Department of Computer Science. I thank AbleLink Technologies for the intellectual collaboration and for providing a solid software foundation for the technology probe. I would like to express gratitude to BEA Systems, especially my manager Scott Musson, for not only providing gainful
v
employment during much of my doctoral work, but for valuing and protecting my academic endeavors and ensuring that I finish with a degree. I acknowledge the many friends who have supported and encouraged me along the way. I thank Anne, my dearest friend and confidant, and the Hendersons – my second family – who have provided so many years of emotional support. I thank Joy and Laura and everyone in the “FAC club” who filled my years in graduate school with adventure and new experiences. I am deeply indebted to Maryanne DeHart (a.k.a Finn Thye), my dear friend and dissertation editor, who painstakingly translated my ideas into words and sentences that are intelligible by others. My deepest thanks go to my research participants, for sharing their stories and lives with me. I especially thank the two probe study families for their incredible commitment and patience throughout this research. Last but surely not least, I thank Brian Dresser for his love and friendship, unwavering support of my work, and daily encouragement even from a distance – reminding me that with lots of deep breaths and many small steps, I would get here.
vi
Contents Chapter 1: Introduction ....................................................................................................... 1 1.1.
Research Problem.......................................................................................... 1
1.2.
Studies and Findings ..................................................................................... 3
1.3.
Reading Guide............................................................................................... 4
Chapter 2: Cognitive Disability and the Untapped Potential of Mobile Communication Technology............................................................................. 6 2.1.
Introduction ................................................................................................... 6
2.2.
Defining Cognitive Disability ....................................................................... 7
2.3.
Supports for Individuals with Cognitive Disabilities .................................... 8
2.4.
Mobile Communication............................................................................... 10
2.5.
Research Goals ............................................................................................ 12
2.6.
Related Work............................................................................................... 13
2.7.
Summary ..................................................................................................... 14
Chapter 3: Research Approach ......................................................................................... 16 3.1.
Introduction ................................................................................................. 16
3.2.
Participatory Design.................................................................................... 16
3.3.
Considerations in My Research Strategy .................................................... 20
3.4.
Cultural and Technology Probes ................................................................. 26
3.5.
Research Approach: Three Research Phases............................................... 28
3.6.
Summary ..................................................................................................... 30
Chapter 4: A View of the Landscape................................................................................ 32 4.1.
Introduction ................................................................................................. 33
4.2.
Methods and Setting.................................................................................... 34
4.3.
Understanding the Adoption Process .......................................................... 37
4.4.
Keep it Small and Simple, Please! .............................................................. 41
4.5.
Hopes, Dreams, and Technology ................................................................ 48
4.6.
Summary ..................................................................................................... 51
vii
Chapter 5: Understanding Remote Communication Today.............................................. 53 5.1.
Introduction ................................................................................................. 53
5.2.
Methods and Setting.................................................................................... 54
5.3.
Nature of Remote Communication ............................................................. 58
5.4.
Challenges with Mobile Phones .................................................................. 61
5.5.
Desire for and Challenges of Social Inclusion ............................................ 64
5.6.
Perspectives on Independence..................................................................... 65
5.7.
Summary ..................................................................................................... 67
Chapter 6: Co-Designing a Remote Communication System........................................... 69 6.1.
Introduction ................................................................................................. 69
6.2.
Research Goals in Phase 3 .......................................................................... 69
6.3.
Adapting the Technology Probe Method .................................................... 70
6.4.
Method and Setting ..................................................................................... 72
6.5.
Participant Families..................................................................................... 74
6.6.
Design Activities before the Technology Probe Use Period....................... 78
6.7.
Activities during the Technology Probe Use Period ................................... 84
6.8.
Daily Use of the Probe ................................................................................ 85
6.9.
Activities after the Technology Probe Use Period ...................................... 89
6.10.
Summary ..................................................................................................... 90
Chapter 7: Implementation, Evolution, and Evaluation of the Probe .............................. 91 7.1.
Introduction ................................................................................................. 91
7.2.
Mobile Technology Platform ...................................................................... 92
7.3.
AbleLink Software ...................................................................................... 94
7.4.
Usage Logging ............................................................................................ 98
7.5.
Data Aggregation and Visualization Tool................................................. 100
7.6.
Linsey’s Probe........................................................................................... 101
7.7.
Margaret’s Probe ....................................................................................... 110
7.8.
Non-Functional Changes........................................................................... 114
7.9.
Challenges of a Mobile Platform .............................................................. 114
7.10.
Design Implications from Usability Findings ........................................... 116 viii
7.11.
Summary ................................................................................................... 117
Chapter 8: Remote Communication Patterns of Two Families ...................................... 118 8.1.
Introduction ............................................................................................... 118
8.2.
Linsey and the Roberts Family: Mobile Phone as Connector ................... 119
8.3.
Margaret and the Smith Family: Mobile Phone as Emancipator .............. 129
8.4.
Summary ................................................................................................... 143
Chapter 9: Social and Emotional Aspects of Assistive Technology.............................. 144 9.1.
Introduction ............................................................................................... 144
9.2.
Informational and Relational Aspects of Family Communication............ 145
9.3.
Relational Aspects of Safety, Independence, and Social Inclusion .......... 149
9.4.
Design Implications................................................................................... 151
9.5.
Summary ................................................................................................... 161
Chapter 10: Design as Reflective Practice: Affordances of an Evolving Probe............ 163 10.1.
Introduction ............................................................................................... 163
10.2.
Benefits and Drawbacks of Technology Probes ....................................... 163
10.3.
Toward an Assistive Technology Design Methodology ........................... 173
10.4.
Understanding Requirements for a Meta-Design Environment ................ 175
10.5.
Summary ................................................................................................... 179
Chapter 11: Summary and Future Work......................................................................... 180 11.1.
Summary of Approach and Results........................................................... 180
11.2.
Future Work .............................................................................................. 183
11.3.
Conclusion................................................................................................. 186
References....................................................................................................................... 187 Appendix A: Phase 1 Interview Questions ..................................................................... 199 Appendix B: Phase 2 Interview Questions ..................................................................... 204 Appendix C: Materials from Phase 3.............................................................................. 210
ix
Appendix D: Glossary..................................................................................................... 213 Appendix E: Abbreviations ............................................................................................ 216
x
List of Tables Table 3.1: Unique aspects of design environment ...................................................... 21 Table 3.2: Three research phases ................................................................................ 30 Table 4.1: Assistive technology devices discussed in interviews............................... 36 Table 5.1: Young adults who participated in Phase 2 interviews ............................... 56 Table 6.1: Participants of the Phase 3 study ............................................................... 74 Table 6.2: Functional description of Linsey's abilities ............................................... 76 Table 6.3: Functional description of Margaret's abilities ........................................... 78 Table 6.4: Design and assessment activities before probe use period ........................ 79 Table 6.5: Activities during the technology probe use period .................................... 85 Table 6.6: Excerpt from the probe log ........................................................................ 86 Table 7.1: Four Categories of Usage Logging............................................................ 99 Table 7.2: Log file entries showing recipient and duration of phone calls............... 100 Table 7.3: Evolution of Linsey's phone .................................................................... 104 Table 7.4: Evolution of Margaret's probe ................................................................. 111 Table 8.1: Linsey's weekly schedule of activities (Monday – Saturday).................. 122 Table 8.2: Purpose of Linsey's phone calls: categories and examples...................... 124 Table 8.3: Margaret's call recipients before and during the probe study .................. 131 Table 8.4: Margaret's weekly schedule of activities ................................................. 136 Table 8.5: Purpose of Margaret's phone calls: categories and examples.................. 139 xi
Table 9.1: Four categories of remote communication between parents and their children, and the informational and socio-emotional purposes of each category ....................................................................................... 147 Table 9.2: Four approaches to incorporating emotion in technology ....................... 155 Table 9.3: Emotional goals observed between care recipients and caregivers ......... 156 Table 9.4: Evaluation framework for emotional support communication................ 158 Table 10.1: Types of meta-design requirements identified by the probe study........ 176 Table 11.1: Summary of research findings and design implications ........................ 182
xii
List of Figures Figure 2.1: Understanding cognitive disability as each person having islands of deficit in seas of ability, as well as islands of ability in seas of deficit ...... 8 Figure 3.1: Design-time and use-time diagram [Fischer, 2000] ................................. 19 Figure 3.2: Challenge of anticipating situated action, and dimensions of evolution.. 25 Figure 3.3: Risk of assuming user motivation before user has experienced value from the system......................................................................................... 25 Figure 4.1: Optimal flow diagram [Csikszentmihalyi, 1990] ..................................... 47 Figure 6.1: Seed – evolutionary growth – reseed (SER) model applied to the remote communication probe study......................................................... 72 Figure 6.2: Linsey Roberts, talking on the phone to her father while on a vacation cruise with her mother Carin (photo blurred to protect privacy) ............. 75 Figure 6.3: Margaret Smith (photo blurred to protect privacy) .................................. 77 Figure 6.4: Sample remote communication tasks with feedback ............................... 80 Figure 6.5: Paper-based mockups for remote communication tasks .......................... 81 Figure 6.6: Annotated color mockup .......................................................................... 83 Figure 6.7: Phone call journal entry of a call Linsey made to her mother.................. 89 Figure 7.1: Verizon XV6700 phone............................................................................ 92 Figure 7.2: Two cases for the phone: (A) a neoprene case designed for an iPod, and (B) a rubberized "flexiskin" case designed for a different phone model............................................................................................. 94 Figure 7.3: Configuration file for customizing the Memo Player application............ 95
xiii
Figure 7.4: AbleLink's software used in the probe study: (A) Pocket Ace, (B) Discovery Desktop, (C) Schedule Assistant....................................... 97 Figure 7.5: Researcher tool developed to view and analyze quantitative and qualitative data collected during the probe study............................ 101 Figure 7.6: Linsey's version of Discovery Desktop (left) and Pocket Ace (right).... 102 Figure 7.7: Linsey’s Memo Player showing the Play screen (A) and the caregiver’s Update Memo screen (B) .................................................... 103 Figure 7.8: Modified "Today" screen with accessible launcher (heart button) ........ 106 Figure 7.9: Screen lock feature showing (A) the unlocked state, (B) Linsey’s version and (C) Margaret's version of the screen lock........................... 107 Figure 7.10: The number pad dialer (A) and morning alarm screen (B) on Margaret’s probe............................................................................. 110 Figure 7.11: Margaret's bus arrival information screen launcher (A) and application screen (B) ................................................................... 112 Figure 7.12: Missed call alert screen that was added to Margaret's probe ............... 113 Figure 7.13: Summary of usability findings through probe study ............................ 116 Figure 8.1: Linsey’s number of calls and their recipients over a 7 day period......... 120 Figure 8.2: Linsey’s call frequency across the day and across days of the week ..... 121 Figure 8.3: Margaret's number of calls and call recipients over total 8 weeks......... 132 Figure 8.4: Two remote communication tasks that were transformed into technology-supported tasks ............................................................ 134 Figure 8.5: Margaret's pattern of probe usage across the study period..................... 135
xiv
Figure 9.1: Diagram of the interconnections and potential tensions between safety, independence, and social connectedness ..................... 150 Figure 10.1: Design-time and use-time diagram modified to include mediated meta-design phase ................................................................ 178 Figure C.1: Phone Diary “cheat sheet” ..................................................................... 210 Figure C.2: Phone call journal template for the Roberts family ............................... 211 Figure C.3: Phone call journal template for the Smith family.................................. 212
xv
A journey of a thousand miles must begin with a single step. – Lao Tsu
1
Introduction Safety, independence, and social inclusion – these three basic goals are fundamental to our human existence individually and as a species. This dissertation explores how technology can play a role in supporting these goals in the lives of individuals with significant cognitive disabilities. More specifically, this dissertation presents an innovative participatory design approach for assistive technologies, and describes findings from applying this approach in the co-design of a socio-technical remote communication system with and for individuals with cognitive disabilities and their caregivers. 1.1. Research Problem Over the past 50 years in the United States, individuals with cognitive disabilities have quietly moved out of institutions and into community-inclusive living environments [Braddock & Parish, 2002]. Due to a grassroots movement by selfadvocates that led to government legislation to increase social inclusion and community participation of these individuals, today many people with cognitive disabilities live independently or in community-based assisted living environments. 1
Most individuals in this population live at home with their families as children and young adults. Many parents of young adults with cognitive disabilities find themselves with new opportunities to involve their children in vocational and community activities; but are also given the added responsibility of acting as their grown child’s primary caregiver and managing their busy schedule, as well as their own. I will refer to this new model for care, where families rather than institutional staff are the primary caregivers, as the “family-based care model.” New tools and technological assistance are needed to support the family-based care model. Assistance is particularly needed during the life transitions that are often more difficult for people with special needs [Dawe, Fischer et al., 2005], such as the transition out of the school system at age 21. Individuals with cognitive disabilities have fewer assistive technology supports than people with other disabilities, even though they have the greatest functional need for technology assistance [Mann, 1993; Riemer-Reiss, 2000]. The lack of assistive technology for this population may be a reflection of the difficulty of creating effective technical supports for cognitive disabilities. The goal of augmenting cognition [Engelbart, 1995] is fundamentally more difficult than augmenting other types of disabilities, such as vision or mobility. The cognitive mechanisms people employ are less understood and less visible than physical mechanisms, and vary according to context, time, and between individuals. Creating a cognitive prosthetic is more challenging than creating a physical prosthetic because of the inherent complexity in determining where cognition “ends” and the prosthetic should “begin” [Newell, Carmichael et al., 2002]. The challenge of creating effective assistive technology is evident in a high abandonment rate: a surprisingly large percentage of assistive technology devices (35% or more) are purchased, but not successfully adopted [Martin, 1999; Riemer-Reiss, 2000; Kintsch & DePaula, 2002]. The lack of effective technology for individuals with cognitive disabilities is also due in part to a shortage of research in empirically-grounded design and evaluation methods for technology for this population. Individuals with cognitive disabilities are severely underrepresented in technology design methods and usability evaluations [Bühler, 2001; Sutcliffe, Fickas et al., 2003; Lewis, 2006]. While it is recognized that
2
design and evaluation methods will need to be adapted for this population [Fischer & Sullivan, 2002; Lepistö & Ovaska, 2004], the paucity of design research in this area makes it difficult for a technology designer to determine how to make appropriate adaptations. 1.2. Studies and Findings The research presented in this dissertation is divided into three phases, each of which comprised a separate study. The first two studies were semi-structured interview studies, and the third study employed multiple methods including participatory design, ethnography, and technology probes. In the first study, I developed a landscape view of assistive technology usage and abandonment among children and young adults with cognitive disabilities. In the second study I explored the strategies and tools families use to support remote communication, and the role remote communication played in their child’s safety, independence, and social connectedness. The third phase was the bulk of the dissertation research. In this phase I co-designed a remote communication system based on a portable digital assistant (PDA), with and for young adults with cognitive disabilities and their families. The remote communication system was iteratively evaluated and modified through a technology probe usage study. Through this research, I identified research and design implications along three different dimensions: •
Technology interface implications. These include design implications based on the users’ interactions with the graphical user interface, as well as the selection of the platform and peripherals.
•
Socio-technical environment implications. These include social and environmental considerations that affect usage.
•
Design methodology implications. These include implications for the methods and conceptual frameworks for assistive technology design.
The findings are presented in context with supporting data in each study, in Chapters 4 through 7. Chapters 8 through 10 elaborate and synthesize these findings with related research. 3
1.3. Reading Guide This dissertation is organized in the following way. I begin by providing an introduction to the research domain and describing my research approach. I then present the methods and data collected chronologically through the three research phases. This is followed by a discussion of the findings and implications for assistive technology design. I conclude with a summary of my contributions and suggestions for future work. In Chapter 2, I introduce the research domain and outline my research goals. I begin by defining cognitive disability, and describe how legislative and societal changes have led to increased community participation by individuals with cognitive disabilities, and increased the prevalence of a family-based care model. I then review recent mobile communication technology research, and describe its largely untapped potential as a vital component of an assistive socio-technical environment for individuals with cognitive disabilities. In Chapter 3, I motivate and define my research approach. I articulate the unique challenges of this research environment, and describe the benefits and challenges of participatory design, meta-design, ethnography, and technology probes as design research methods for this environment. I then present my three-phased research approach. In Chapter 4, I present the method, data and findings from the first research phase, which provides a view of the landscape of assistive technology adoption by families who have a child with cognitive disabilities. I describe the prevalence of the caregiver network in the adoption process, the need for simple devices that are easy to troubleshoot and maintain, and parents’ hopes and goals for assistive technology to increase their children’s independence and social inclusion. In Chapter 5, I present the method, data and findings from the second research phase, which showed me families’ current methods of remote communication. I describe usability challenges with existing mobile phones and the role of remote communication in safety, independence, and social inclusion. Chapters 6, 7 and 8 are dedicated to the third research phase, the technology probe study. In Chapter 6 I describe the method of the study, including the design and 4
research activities before and during the probe usage period. Chapter 7 presents the probe implementation, including how the probe evolved over the course of the usage period. In Chapter 8, I present the data collected and an analysis of the usage patterns within the socio-technical environment. In Chapter 9, I synthesize the findings across the three phases to demonstrate that emotional and relational needs are central to the family-based care model. I provide a summary of existing emotion research in cognitive science and human-computer interaction (HCI), and argue that design researchers need to focus on users’ relational and emotional goals. In Chapter 10, I analyze the effectiveness of the research methodology itself. I discuss the affordances of a technology probe, including its facility in engaging individuals with cognitive disabilities and their parents as active participants in the design process. I suggest that this method informs requirements of a meta-design environment through a mediated meta-design process. In Chapter 11, I conclude this dissertation with a summary of the research and a discussion of opportunities for future work.
5
Accept us for who we are – without any reservations. With your help and encouragement, we can become responsible and contributing members of our communities. – Annie Forts, Self-Advocate, “Up” Syndrome Fund
2
Cognitive Disability and the Untapped Potential of Mobile Communication Technology In this chapter I provide a brief background of the domain of this research. I define cognitive disability and attempt to summarize the social issues that surround young adults with significant cognitive disabilities and their families in the United States today. I introduce mobile communication technology and the major social impact that widespread mobile phone adoption has made on our society. I discuss research that suggests the potential for this technology to improve the lives of individuals with cognitive disabilities. Finally, I articulate my research goals and review related work in this area. 2.1. Introduction This research took place in the homes and communities of families with a young adult son or daughter with cognitive disabilities. Like all families, each of these families is a complex social system made up of close interdependent relationships. Unlike many families, these families must overcome daily obstacles in societal acceptance, stereotypes, and ongoing battles for much-needed government support. It
6
is no wonder that many of the parents in these families dedicate their lives to advocacy. The families I have worked with in this research have all been amazingly committed to improving the life of their child with cognitive disabilities. Research in this setting means working with participants who are extremely busy, yet very dedicated to participating in efforts to improve the quality of life for their child. 2.2. Defining Cognitive Disability Over 20 million people living in the United States have a cognitive disability, which includes intellectual disabilities (also referred to as developmental disabilities), traumatic brain injury, Alzheimer’s disease, and sometimes stroke [Braddock, Rizzolo et al., 2004]. Intellectual disabilities have a variety of developmental etiologies including Down syndrome, certain genetic disorders, and birth defects. Some individuals with cerebral palsy and autism spectrum disorder also have intellectual disabilities. Individuals with a cognitive disability have substantial difficulty with thought processes. This includes conceptualizing and understanding abstraction, planning and sequencing, understanding numbers and other symbols, and interpreting subtle social cues [Braddock, Rizzolo et al., 2004]. These difficulties impact every aspect of one’s life. According to the Diagnostic and Statistical Manual of Mental Disorders [American Psychiatric Association, 2000], a person with cognitive disabilities is “significantly limited in at least two of the following areas: self-care, communication, home living social/interpersonal skills, self-direction, use of community resources, functional academic skills, work, leisure, health and safety.” It is very difficult to generalize the needs and abilities among individuals with cognitive disabilities. Even within a single category such as Down syndrome or autism, each individual has a unique profile of skills and needs [Bird & Buckley, 1999]. This can be viewed as a problem of designing for a universe of one [Fischer & Sullivan, 2002], a term first used by psychoanalyst Erik Erikson to describe the uniqueness of each patient’s case (as described by Schön [1983]). Abilities differ across individuals and for a single individual across contexts and time periods. Individuals will surprise you with unexpected abilities and disabilities. Elliot Cole, a
7
senior researcher in the assistive technology field, expressed this through a tangible metaphor: each individual has unique islands of abilities in seas of deficits, and islands of deficits in seas of abilities, shown in Figure 2.1 [Cole, 2006].
Figure 2.1: Understanding cognitive disability as each person having islands of deficit in seas of ability, as well as islands of ability in seas of deficit Having a significant cognitive disability used to mean that one would live their life in an institution, away from family and society. Until 30 – 40 years ago, most individuals with moderate to severe cognitive disabilities were tucked away in large institutions with very little opportunity to contribute to the broader community, and were effectively hidden from mainstream society [Braddock, 2002]. Through a social movement supported by federal legislation, individuals with disabilities have increasing opportunities for social inclusion and community participation. Children with cognitive disabilities are increasingly able to attend mainstream public schools. Most live at home with their families as children and young adults, and eventually move into a community-based assisted living environment. In 1990 the federal government passed the Americans with Disabilities Act (ADA), which encourages integration and prohibits discrimination against individuals with cognitive disabilities in areas including employment, housing, transportation, recreation, health services, and access to public services [ADA, 1990]. 2.3. Supports for Individuals with Cognitive Disabilities While in the school system, children with cognitive disabilities receive support from a network of professional caregivers, including special education teachers, paraprofessionals (teachers’ aids), speech therapists, physical therapists, assistive technology specialists, and others. Many students with special needs also have access to a wide variety of assistive technology owned by the school system. 8
2.3.1. Assistive Technology Assistive technology (AT) is defined by the ADA as “any item, piece of equipment, product, or system, whether acquired commercially or off the shelf, modified, or customized, that is used to increase, maintain, or improve functional capabilities of individuals with disabilities” [US technology-related assistance for individuals with disabilities act, 1988]. Assistive technology ranges from low-tech devices such as canes and magnifying lenses to sophisticated computer software and hardware. For individuals with cognitive disabilities, assistive technology includes augmentative communication devices (informally called “talkers”) for nonverbal individuals, word prediction and spelling software, screen reading software, and alternative input devices (e.g., simplified keyboard displays) [Dawe, 2006]. 2.3.2. Transitioning from the School System Unfortunately, much of these supports drop off, and the accumulated knowledge about an individual is lost, when the individual leaves the school system, [Dawe, Fischer et al., 2005; Dawe, 2006]. When their children turn 21, many parents of young adults with cognitive disabilities find themselves with the new responsibility of being their grown child’s primary caregiver. While families continue to receive government support, the resources that scaffolded the child’s education are no longer present. Parents are responsible for managing their child’s busy schedule (which often includes a job and various community activities), as well as providing transportation and coordinating with other caregivers. Most parents are not yet at retirement age during this transition, so they are also managing their own work and private lives. These parents have some support by a government-provided case manager, but in large part must develop their own strategies for providing care. The support for finding, purchasing, and configuring assistive technology drastically declines after the child leaves the school system, so adults with cognitive disabilities use far fewer assistive technology devices [Dawe, Fischer et al., 2005]. New tools are needed to support this family-based care model, such that parents feel secure and confident letting their children go out into the community, while
9
individuals with cognitive disabilities safely reach their potential as active community participants. 2.4. Mobile Communication Mobile communication technology has transformed our society, including the way we work, play, keep in touch with loved ones, and organize our lives [Ling & Yttri, 2002; Palen, 2002; Lindholm, Keinonen et al., 2003]. These lifestyle changes may include staying in “perpetual contact” and increasing our connectedness with our social network of friends and family wherever we are [Katz & Aakhus, 2002], relying on real-time coordination and organization of social activities instead of carefully planning ahead of time, and acting with an increased sense of security because of the ability to call someone at any moment if we are in need of help [Ling, 2004]. The way that mobile phones are used and perceived varies by individual, but there appear to be trends of types of usage based on social surroundings and age. Ling and Yttri [Ling & Yttri, 2002] identified three categories of functions or roles of mobile communication in society: security, coordination, and expressiveness (social interaction). They found that teens primarily use the mobile phone for expressiveness, to interact socially and emotionally with their social network, and to maintain and improve their social position within the network. Active adults, on the other hand, mainly use their mobile phones for coordination of daily life, often making and changing plans in a fluid, dynamic way based on changing conditions in the environment; this rapid, frequent contact for planning is termed micro-coordination. Finally, the elderly viewed security as the primary reason to adopt and use a mobile phone. A number of studies have shown a special role that mobile phones play in the lives of individuals who are in chronic need of help or who provide care to others [Katz & Aakhus, 2002; Ling, 2004]. Those in chronic need of help include the elderly, the medically fragile, and individuals with disabilities. Ling [2004] reports on a number of studies that demonstrate mobile phones acting as a lifeline between individuals who need help, and their caregivers who want to make sure they are always in reach to provide help. Even for those who do not often need to use them,
10
mobiles increase peoples’ perception of security by providing insurance against potential dangers or difficulties. In one example, elderly parents explained that they only use their mobile phone during trips into the mountains, when they call their adult children to let them know where they are going and for how long [Ling, 2004]. In Israel, there is a mobile phone service plan that restricts the outgoing calls to only a few numbers, designed specifically for safety and monitoring of the elderly and children [Schejter & Cohen, 2002]. Mobile phones have been shown to give increased freedom and independence to mobile users, and in particular users with disabilities who are often underserved by technology. Palen [2002] describes a case of a man in a wheelchair who experiences new freedom with a mobile phone, as he becomes able to spend time and work alone on his property, without the worry that he might fall out of his wheelchair. Another population that has benefited from mobile phones are the deaf and hearing impaired who use text messaging, which reduces their dependence on specialized hardware for remote communication [Power & Power, 2004]. 2.4.1. User-Interface Challenges for Individuals with Cognitive Disabilities Although the elderly and people with disabilities have seen benefits from mobile phones, these devices are generally not designed with these groups in mind. Consequently, the interfaces of off-the-shelf designs impose serious usability difficulties and do not support these users’ specific needs. Standard HCI issues, including a complex user interface (UI), tiny buttons, and a complex navigation system make phone use difficult for a person who is unable to read or does not have fine motor control. In general, mobile phones are becoming smaller and lighter, while also packing more functionality into a reduced screen. Many mobile phones come with dozens of features including text messaging, a digital camera, calendars, games, even music and video players that complicate the interface. For novice mobile phone users and non-traditional user populations like the elderly, this is moving in the wrong direction. Recent studies with elderly mobile phone users [Osman, Maguire et al., 2003; Ziefle & Bay, 2004] have shown the need for simpler navigation and menus,
11
text that is clear and easy to read, and larger buttons. These users are willing to give up some of the functionality for simplicity. 2.4.2. Socio-Technical Challenges for Individuals with Cognitive Disabilities More fundamentally, little research has been done to understand how mobile phones might be designed to serve those with special needs, and in particular, to help people with cognitive impairments. For example, increasing security for this population may mean that if the caregiver does not respond to a call, the caller is immediately forwarded to a 24 hour care service. Recently a few advances have appeared in this area, such as simpler handsets like the “Migo” [Verizon Wireless, 2006] and child location services for parents by wireless providers. On the research side, a recent survey and field study with youths (ages 7 to 17) with cognitive impairments and their teachers and parents explored how they use PDAs and mobile phones [Gillette & DePompei, 2004]. This study found that under 35% of the participants used mobile phones, and even fewer used more targeted mobile devices such as electronic organizers. They suggested this was because there have been few studies to match features and functions of devices to the needs and abilities of these youths. Additional research is needed to understand the role(s) that mobile phones can play in the lives of individuals with cognitive disabilities and in aiding their caregivers in providing care. Mobile communication has the potential to bring increased independence and greater access to social opportunities, which includes both the ability to communicate with people who are not physically present and to plan and coordinate future face-to-face activities while mobile. 2.5. Research Goals The objectives of this research are to contribute to the methods, empirical knowledge, and theoretical issues involved in designing remote communication technology with and for individuals with cognitive disabilities and their families. Specifically, my research goals are:
12
•
To identify and evaluate methods for developing technology for individuals with cognitive disabilities in which caregivers and care recipients are active participants in the design process;
•
To develop an empirically-grounded understanding of the socio-technical nature of remote communication between individuals with cognitive disabilities and their caregivers, and to explore the role of such communication in safety, social connectedness, and independence;
•
To identify and understand the major usability issues of using mainstream PDA devices as mobile assistive devices, through an extended field study with individuals with cognitive disabilities; and
•
To advance theoretical and practical understanding of meta-design by designing, building, and evaluating a system to support the meta-design principle of seeding, evolution, and reseeding at “use-time.”
2.6. Related Work The past few years have seen an explosion in field research in mobile technology and particularly mobile phones, including [Pascoe, 2000; Perry, 2001; Palen, 2002; Schejter & Cohen, 2002; Lindholm, Keinonen et al., 2003; Newcomb, Pashley et al., 2003; Holtzblatt, Ketola et al., 2004; Hulkko, Mattelmaki et al., 2004; Ketola, 2004; Ling, 2004]. While researchers recognize the new challenges to studying technology usage on a small device in a mobile context [Holtzblatt, Ketola et al., 2004], it is widely held that field research is vital to understand how these mobile devices are incorporated into and transforming our lives [Nieminem-Sundell & VaananenVainio-Mattila, 2003]. Most field studies with mobile devices have been in the context of computer supported collaborative work (CSCW), where researchers study site-independent mobile workers [Pascoe, 2000; Perry, 2001; Hulkko, Mattelmaki et al., 2004] or people engaged in a very specific task like grocery shopping [Newcomb, Pashley et al., 2003]. Fewer but increasing studies describe the use of mobile devices in the home or for leisure activities [Palen, 2002; Schejter & Cohen, 2002; Holtzblatt, Ketola et al., 2004; Ling, 2004].
13
In studies of the social role of mobile communication, the effect of increased social connectedness is often recognized [Ling & Yttri, 2002; Palen, 2002; Ling, 2004]; however, mobile phones are often not the focus in this research. Recently, researchers have become interested in ways to use ambient and unobtrusive technologies to increase social connectedness between extended or separated families [Hutchinson, 2003; Van der Hoog, Keller et al., 2004], between the elderly and family caregivers [Mynatt, Rowan et al., 2001; Hutchinson, 2003; Morris, Lundell et al., 2004; Van der Hoog, Keller et al., 2004], and between individuals and their community. Many of these projects focus on the elderly. For example, Gaver’s [Gaver, Dunne et al., 1999] Presence Project explores creative ways to increase the social presence of elderly in the community. Researchers at Intel are developing ambient technology to support elders coping with cognitive decline and their extended families [Morris, Lundell et al., 2004]. One very relevant finding from their work is that elderly parents do not want to feel like they are being monitored, but want to have a reciprocal relationship in which they are sharing and contributing as well. A small body of work has emerged in the last few years studying ways to apply user-centered design methods for technology for people with disabilities, including cognitive decline [Osman, Maguire et al., 2003; Morris, Lundell et al., 2004; Schulze, 2004; Newell, Morgan et al., 2006], traumatic brain injury (TBI) [Davies, Marcella et al., 2004; Schulze, 2004; Fickas, Robinson et al., 2005; Todis, Sohlberg et al., 2005], and cognitive disabilities [Rousseau, 1999; Carmien, 2003; Crabtree, Hemmings et al., 2003]. Researchers recognize both the need for increased sensitivity in these environments, and the added challenges that are faced by the designers in applying traditional software design methods. A qualitative or case study approach is often used to gather in-depth and detailed information about potential users, which can help overcome designers’ biases towards the ‘typical’ user. 2.7. Summary In this chapter I provided a glimpse into the lives of individuals with cognitive disabilities and their families. I articulated the need for new supports for young adults
14
with cognitive disabilities and their parental caregivers, in light of the recent shift from an institutional to a family care model in the United States. I then described the major impact that mobile communication has made in mainstream society by supporting security, coordination, and expressiveness (social interaction) in new ways. I discussed the UI-level and socio-technical level issues presented by off-the-shelf mobile phones for individuals with cognitive disabilities. At the end of this chapter I articulated the major objectives of this research in terms of methodological, empirical, and theoretical contributions. I finished with a discussion of related work in this area.
15
One should never impose one's views on a problem; one should rather study it, and in time a solution will reveal itself. – Albert Einstein
3
Research Approach 3.1. Introduction In this chapter I describe my research strategy, including my research methods and motivations for choosing them. I begin by discussing participatory design and how it relates to traditional software design approaches. I then describe three ways my research context differs from typical software design projects. I then discuss the challenges and importance of designing with individuals with cognitive disabilities, and how I chose a combination of methods from ethnography, participatory design, and a newer method called technology probes. In the subsequent chapters I describe each of my three research phases in more depth. 3.2. Participatory Design Participatory design is predicated on the belief that end-users of technology should be included as participants in the design process. In the fields of HCI and computer-supported cooperative work, this methodology and the philosophy behind it have greatly influenced approaches to including users in the software development process [Carroll, 2006]. 16
3.2.1. A Brief (Selective) History of Software Design Methodologies Software design began as end-user design: computer professionals programmed for their own use, or for the needs of individuals in the same domain of expertise. Computer users and computer professionals were typically one and the same. As computers became more prevalent and moved into businesses and homes, software design increasingly became a process of designing systems for others, and in domains where computer professionals had little knowledge. This section will review various design methodologies that have emerged, in order of increasing level of user involvement. 3.2.1.1. Professional Design Professional design is a term to describe software design that stays completely in the realm of computer professionals, with little or no end-user involvement. Many (if not most) software companies today use this approach. Software engineers (programmers) design and write the software. Then, the same engineers create the UI, or if they are lucky a human factors specialist designs it for them. The goals of the software are determined by engineers or other members of the project, such as product managers, executives, and marketing professionals. The software functionality is based on specifications that sometimes include user goals, such as use cases [Cockburn, 2001], which are fictional representations of users created by members of the project team. The first involvement of users is when the software is released to the market and users buy it (or do not buy it!). 3.2.1.2. User-Centered Design Don Norman [1988] describes User-Centered Design as a new approach that focuses on the user during the design of software systems. The philosophy of usercentered design is that users must be studied so that their needs and abilities are understood and considered in the design of software systems. User-Centered Design has now become a general term for a general orientation to the user. It is not a specific methodology, but implies that the design team is collecting data about users and analyzing this data to create a design that will support and simplify user tasks.
17
3.2.2. Participatory Design User-centered design focuses on the user, but assumes that user needs can be collected and sufficiently communicated to a professional design team, who will use them to design an appropriate solution. In contrast, participatory design [Kensing & Munk-Madsen, 1993; Muller & Kuhn, 1993; Schuler & Namioka, 1993] philosophy asserts that users must become designers themselves, and be actively involved in the design process as members of the design team. Designers are likewise encouraged to spend an extended amount of time at the users’ place of work, both learning and participating in relevant contexts and tasks. Participatory design is rooted in the philosophy of Scandinavian trade unions that asserted that workers should be empowered with new technology, not disenfranchised and replaced by it. Similar to user-centered design, participatory design is a collection of methods and approaches rather than a single methodology. Methods include design workshops, role-playing scenarios, storyboards, and ethnographic methods [Nardi, 1997] such as interviews and participant observation. 3.2.3. Beyond Participatory Design: Towards a Meta-Design Approach One of the criticisms of participatory design is that it takes place entirely at “design-time” (before the system is used), while many of the most important design problems only emerge during “use-time.” This is represented in the diagram in Figure 3.1 [Fischer, 2000]. In traditional participatory design, users or user representatives participate in design activities before use begins.
18
Figure 3.1: Design-time and use-time diagram [Fischer, 2000] Like participatory design, meta-design emphasizes the empowerment of users as active designers, rather than as passive consumers of the technological systems they use [Wright, Marlino et al., 2002; Fischer, Giaccardi et al., 2004; Fischer & Giaccardi, 2006]. Meta-design asserts that system designers cannot know in advance all the situational contexts [Suchman, 1987] in which the system will be used, so the system must be designed to be flexible and open so that it can be extended during use-time. This approach assumes that the very existence of new technology will create emergent needs as users, tasks, and the social context change as a result of the technology. Meta-design transcends participatory design by asserting that the role of user-as-creative-designer must continue to be supported in use-time. The meta-design approach is grounded in a model supporting the ongoing evolution of technological systems through a process of seeding, evolutionary growth, and reseeding (SER) [Fischer & Giaccardi, 2006]. In the beginning, systems are not finalized and closed products, but instead are intentionally underdesigned and act as seeds for ongoing change through contributions by many users. Underdesign does not imply less effort or forethought, but on the contrary it means that the system is carefully designed to be open and extended by end-users. The evolution of the system takes place through periods of incremental, evolutionary growth when users are interacting with the system and making small changes. This is punctuated by episodes
19
of reseeding when designers and users come together and actively re-organize the system to integrate the evolutionary changes. In summary, meta-design encompasses the following themes: •
Design is a co-creative process between technologists and end-users
•
Designers cannot predict how the system will be used, so end-users must be empowered to act as designers during use-time
•
Socio-technical systems will and should evolve over time; a model to support this is the seed-evolutionary growth-reseed (SER) process
•
Technological systems should initially be underdesigned, in the sense that they are not finished products, but rather are flexible design spaces that can support multiple solutions
3.3. Considerations in My Research Strategy In developing an appropriate research approach, I needed to consider the unique challenges of designing technology with and for individuals with cognitive disabilities. I will now describe these challenges, and how I have chosen a combination of design methods to address them. 3.3.1. Unique Aspects of Designing with Individuals with Cognitive Disabilities There are three ways this design environment differs from a typical software design project. These are summarized in Table 3.1. First, individuals with cognitive disabilities have a wider range of cognitive and physical abilities, and each individual represents a “universe of one,” as described in Chapter 2. As a result, designers can make fewer assumptions about users’ abilities and ways of interacting with technology [Newell, Carmichael et al., 2002]. Lightweight user-centered design methods like single interviews and brief observations are not sufficient to develop a good understanding of a person’s skills and practices. Basic user models that underlie many UI evaluation techniques must be reconsidered for users with disabilities [Keates, Clarkson et al., 2000]. This is further complicated because the user is challenged in describing his or her own situation and motivations. Often, the skills
20
central to abstract conceptualization and reasoning are very limited, and users may be unable to conceptualize and verbalize their individual needs and preferences. Unique Feature of Design Environment
Design Implication(s)
Users have widely varying abilities and needs (“universe of one”)
• Lightweight design methods and usability evaluations that are based on typical user behavior are less effective
• User models must be re-evaluated Users are represented by themselves and a network of caregivers
• Caregivers must be included in requirements gathering and design
• Design and evaluation must account for the goals and needs expressed by users and the network of caregivers
Family home and community setting
• Workplace observation methods are not appropriate because of the private home setting
• Environment is less predictable and stable when users go into community
Table 3.1: Unique aspects of design environment A second way this environment differs from a typical user-centered design scenario is that the user is represented by herself as well as through a network of caregivers, among whom there is distributed knowledge about the user’s abilities, interests, and behaviors [Carmien, Dawe et al., 2004]. Caregivers also have hopes and goals for the individual with cognitive disabilities, and the motivations of the caregivers must be understood and balanced with the individual’s motivations during the design process. A third consideration for my research is that the setting is in family homes and communities, in addition to the workplace. The focus of this project is assistive technology, so the research must take place everywhere that users go – it cannot be constrained to a specific location or context. Conducting design in the home environment introduces dimensions of accessibility, privacy, dynamic and ad-hoc organization, and a more delicate environment that changes with the presence of the researcher [Crabtree, Hemmings et al., 2003; Hutchinson, 2003; Bennett, Lindgaard et al., 2006]. 3.3.2. Importance of Including Individuals with Cognitive Disabilities in Design Involving individuals with significant cognitive disabilities in traditional participatory design activities is challenging for various reasons, including limited 21
language ability to describe their needs and opinions, and difficulty participating in abstract or hypothetical discussions and scenarios. Some researchers address this challenge by employing caregivers as “proxy users” to represent the needs of individuals with disabilities [Boyd-Graber, Nikolova et al., 2006]. It is essential to involve caregivers as designers in assistive technology because they are an important source of knowledge and very often will be responsible for configuration and some customization of the technology. However, the “proxy user” approach is problematic because it rests on the assumption that caregivers have perfect knowledge about their care recipients. Design methods relying on designer or caregiver introspection, or formal user models, are less effective with this population [Keates, Clarkson et al., 2000] because these users have a wider range of abilities and needs. Research has found that caregivers do not always have perfect knowledge of the abilities, behavior, and knowledge of their care recipients [Carmien, 2006], nor do they usually have extensive expertise in technology to predict how it should be applied. Indeed, I found that parents were often not able to articulate how technology could be used to increase independence for their son or daughter, although many felt confident that technology would play a key role in doing so. I also found that the goals, motivations, and desires of individuals with cognitive disabilities do not always match those of their caregivers. For these reasons, directly including individuals with cognitive disabilities in the design of their own technology is crucial. 3.3.3. Participatory Design and Ethnography Benefits and Limitations For the reasons outlined above, a carefully crafted design approach that considers these dimensions is crucial. Ethnography and participatory design are well-known methods for incorporating the usage environment and context in design consideration. Participatory design can make important contributions to the methodology to ensure that users are co-creators of their own technologies; however, the methods must be adapted for an environment where caregivers and users are vested stakeholders, and have very different ways of contributing to the design process. One challenge with participatory design is the “tunneling problem,” in which users will only create the patterns of activity to which they are already accustomed, and are unable to transcend
22
their current practices [Crabtree, 1998]. Another limitation of participatory design is that it only considers the technological system during the design phase, and does not address how the system must change and evolve during use-time [Fischer, Giaccardi et al., 2004]. Ethnographic methods such as in-depth semi-structured interviews, participant observation, and diary studies [Berg, 2004] provide the designer with rich real-world data about users’ social and cultural environment. Yet there are challenges to incorporating ethnographic methods into this design environment. A lengthy semistructured interview may not be appropriate with a person with cognitive disabilities. The family home setting makes it more difficult to conduct participatory observation, since a researcher cannot unobtrusively observe family activities in the home as one can observe in a public space or workplace. Another issue is the inherent tension between the ethnographer’s traditional role as an observer, and the designer’s role as a change agent [Button & Dourish, 1996; Crabtree, 1998; Crabtree, 2004]. Even if this is perhaps more of a philosophical problem than a practical one, there is the very practical problem in data collection. Ethnographers frequently spend years collecting sufficient data to describe a social situation; during this time they dedicate their lives to the project and immerse themselves in the culture of study. Technology designers almost never have the time nor the inclination for such an undertaking. I combine traditional ethnographic methods with a technology probe, described below in Section 3.4, to mitigate some of these challenges. 3.3.4. Meta-Design Benefits and Challenges The meta-design approach recognizes that a technological system will need to change and evolve during use-time according to the unique needs of different users. Meta-design is relevant in my design environment because it can help address the following issues: •
Wide variations in user ability. The UI and functionality of the device will need to be customized because of the variance in individual users’ abilities. In the family-based care model, this will most likely be by a family caregiver.
23
•
Frequent changes in communication needs. Parents frequently update their children’s augmentative communication devices based on changing activities and the need for specific content, due to limited abstraction capabilities (e.g., to talk about what the individual did last weekend, or refer to what is for dinner that night). This implies that the remote communication device may also need to be frequently updated.
•
Changing usage environment. The environment in which remote communication takes place is far more dynamic than a workplace environment, and changes in users’ needs and abilities and in the environment will likely introduce new requirements on supporting technology.
Figures 3.2 and 3.3 illustrate two challenges encountered when designing a metadesign environment with new technology. First, there is a bootstrapping problem in anticipating the nature of evolution [Suchman, 1987] before the user has begun to use the technology. Designers need to anticipate which system modifications should be possible, and which should be easy. Since every configuration option and UI element adds complexity, each dimension of customization must be considered judiciously. Easy modifications should be in areas of the system where the desired behavior and or appearance of the system are very likely to differ between users, and should be supported as customizations before or during use-time. Possible (but less easy) modifications can be in areas where the behavior or appearance of the system are likely to be the same for most users, and can be treated as more advanced capabilities. Some aspects of the system need to be made intentionally difficult to modify to prevent accidental breakage of core functionality. A second challenge encountered in meta-design is that before use, new envisioned technology will have limited meaningfulness for users, resulting in a limited motivation to participate in design. Even when users are highly motivated, most are unable to predict and articulate their own contextual behavior before they have incorporated the technology into their lives.
24
Figure 3.2: Challenge of anticipating situated action, and dimensions of evolution
Figure 3.3: Risk of assuming user motivation before user has experienced value from the system In my design approach, I mediate these challenges by creating a design-in-use process. I combine participatory design with technology probes, which incorporate technology usage as part of the design process.
25
3.4. Cultural and Technology Probes Cultural and technology probes [Gaver, 2002; Hutchinson, 2003] have recently emerged as a potential nexus between ethnographic research and participatory design. Gaver [Gaver, Dunne et al., 1999] developed the concept of cultural probes as a way to learn about users and inspire them to develop new ideas. Cultural probes are “kits” of simple objects that are given to participants to collect ethnographic information about their own lives. Cultural probe kits usually include evocative and fun activities that involve a disposable camera, a diary, a photo album, and other simple objects. The data collected through these objects can then be used in workshops with designers and users to inspire new technology. Cultural probes have been used in projects where computing is taken out of the workplace and with non-traditional user groups such as families [Hutchinson, 2003], the elderly [Gaver, Dunne et al., 1999] and the cognitively impaired [Crabtree, Hemmings et al., 2003]. Technology probes [Hutchinson, 2003] were inspired by cultural probes as a way to field test new technology and gather design requirements at the same time. Technology probes are not simply a digital version of cultural probes; whereas cultural probes are employed before the design phase to inspire the design of future technologies, technology probes are used during the design phase to gather data about potential technology usage, enrich the design space, and inspire users to think about new ways to use technology. In addition to information collection, technology probes provide practical functionality to the user. Technology probes are digital artifacts that are designed to support the behavior that the researcher is interested in understanding; during use a probe quietly “observes” user behavior by recording extensive usage data. This data is then analyzed in concert with interview and human observation data. Rather than assuming the stance of a passive ethnographic observer, technology probes intentionally enrich the design environment, changing the way users engage in activities and encouraging users to think about new ways to leverage technology in their lives.
26
3.4.1. Contexts of Use Technology probes have been used in design environments where it is difficult to apply standard ethnographic methods, such as in the family home [Hutchinson, 2003] and during highly mobile activities [Hulkko, Mattelmaki et al., 2004]. For example, researchers in the interLiving project [Hutchinson, 2003] sought to understand the nature of communication between extended family members, particularly across generations. The researchers used a technology probe that was a touchscreen tablet with a shared display, placed in each family home, providing family members across homes the ability to write memos to one another. In a second probe study, the probes were shared digital picture frames that allowed easy creation and sharing of images between adult siblings who lived apart. Technology probes have been used to collect data in highly mobile contexts, such as activities of shoppers and “site-independent mobile workers” [Hulkko, Mattelmaki et al., 2004]. In these studies technology probes were based on mobile phones, and behaved like digital diaries [Masten & Plowman, 2003] in that users recorded data about their activities on the phone. In the mobile worker study the probe actively solicited information by sending users specific, contextual questions via text messages (also known as short message service (SMS)). 3.4.2. Designing a Technology Probe A technology probe must be designed with four properties in mind. First, it should provide very simple functionality that is easy to use, and stable enough to use for an extended field study. Second, it should support user activities in a domain that designers are interested in understanding. Third, it should be instrumented to detect and log extensive data during use, so that designers can reconstruct usage. Last, it should be built on technology from which designers want to gather usability data. If designed successfully, a technology probe can provide the following benefits [Hutchinson, 2003]: •
Ethnographic benefit: understanding the needs and attitudes of users in a real-world setting
27
•
Evaluative benefit: gathering usability data about new technology in the field with users from your intended user population
•
Design benefit: inspiring users and researchers to think about uses of new technologies during use-time
Probes are not an entirely new idea in the field of HCI. Probes are related and influenced by diary studies [Rieman, 1996; Brown, Sellen et al., 2000; Palen & Salzman, 2002], which have been used successfully in environments that are difficult to get access to, such as the home. Two differences between diary studies and technology probes are: 1) probes change the design environment by introducing technology that supports and inspires users to do things in new ways, whereas diary studies do not significantly change the environment and are meant to collect data about existing practices; 2) probes can collect usage data unobtrusively, in addition to actively soliciting user feedback, versus diary studies which strictly collect selfreported user information. Technology probes are also similar to prototypes and instrumented applications (e.g., see [Todis, Sohlberg et al., 2005]), both well-known software design tools. A technology probe may sometimes be a prototype of a future system. However the primary function of a technology probe is to inspire designers and users to learn and reflect on user practices and the potential of technology in the design of future technologies. A technology probe is an example of a certain type of instrumented application whose data is used for ethnographic, evaluative, and design purposes. The category of “instrumented applications” is broad and does not specify how the collected data is used, and often refers to applications that gather usage data in production for performance tuning. 3.5. Research Approach: Three Research Phases This research project consisted of three phases, where each phase was grounded in and progressively built on the previous phase. Through this phased approach, I first developed an understanding of how things are by studying current practices and technologies used and then moved to exploring how things might be by introducing new technology into the environment. 28
The first phase was a semi-structured interview study with parents and teachers of students with cognitive disabilities. The purpose of this phase was to gain a broad understanding of the role that assistive technology plays today for these groups, and identify the benefits of and barriers to assistive technology adoption and use. I learned about current technology that caregivers are using to provide care, and technology young adults with cognitive disabilities are using (including assistive technology and general technology such as video games and portable music players). I explored issues and barriers in assistive technology adoption that parents and teachers have experienced. Finally, I asked parents and teachers to share their hopes and dreams about assistive technology so that my research could be guided and inspired by the visions of my potential users. In the second phase, I took a closer look at a few themes that emerged in Phase 1: remote communication, independence, and social connectedness. Phase 2 was a more in-depth semi-structured interview study with fewer participants. The research setting was the family home. I interviewed parents and their young adult children with cognitive disabilities to learn about ways that caregivers and care recipients achieve remote communication today, and the role that it plays in increasing independence and safety. I explored perceptions of independence and safety, among parental caregivers as well as young adults with cognitive disabilities. While some of the young adults with cognitive disabilities used mobile phones and some did not, I identified common themes in requirements and desires for an accessible mobilephone based remote communication system. In the third phase, I conducted an evolving technology probe study with two of the participant families from Phase 2. The purpose of this phase was to understand how a handheld remote communication system can support the remote communication tasks specific to each family, and how each family’s needs and requirements change over time through realistic use. In this phase I co-designed the technology probe through participatory design activities with each family. The functionality and user interface reflected the needs and abilities of the child with cognitive disabilities and the goals and practices of the families. The technology probe simultaneously supported remote communication and unobtrusively “observed”
29
the user in communication tasks. In probe fashion, I also added features to the system meant to inspire and provoke families to reflect on ideas about future technology. During this phase data was collected through probe usage logging, a nightly voicemail diary, semi-structured interviews and observations. I conducted participatory design activities during the probe study to modify and evolve the probe, implemented the modifications on the participants’ system, and assessed the modifications through usability testing and naturalistic use. Evolution was based on usability problems as well as new ideas that emerged through use. The three research phases are summarized in Table 3.2: Phase
Method
Participants
Duration
Information Collected
Phase 1
Interview study
Teachers, parents, teenagers and young adults with cognitive disabilities (n=20)
1-2 hour interviews
• Current and previous
Parents and young adults with cognitive disabilities (n=5)
3-4 hour interviews (usually over two meetings)
Phase 2
Interview study
technology used
• Successes and barriers in technology adoption
• Hopes and dreams of families of future technology
• Day-to-day activities of family, focusing on remote communication tasks between caregiver and recipient
• Perceptions of independence and safety among caregivers and care recipients
Phase 3
• Participatory design activities
• Evolving technology probe study
• Diary study • Interviews and
Parents, nonfamily caregivers, young adults with cognitive disabilities (n=2)
10 weeks
• Extensive usage data collected by technology probe
• Nightly diary data reporting usage, problems, new ideas
• Ethnographic data on evolving usage and needs with the probe
observations
Table 3.2: Three research phases 3.6. Summary In this chapter I give an overview of the research approach used in this study. I describe the unique design environment of designing assistive technology for individuals with cognitive disabilities. I motivate the need to include individuals with cognitive disabilities into the design process, and why this requires a carefully crafted 30
design approach. I describe how I have drawn from various research and design methods including ethnography, participatory design, meta-design, and probes, to create a research methodology that explores how things are and how things can be. The research is divided into three phases, which together explore the needs and desires of families who have a young adult child with cognitive disabilities. My research includes working with two families as co-designers of a new assistive remote communication system. In the following chapters, I describe the method, participants, data, and analysis of each of these phases in more detail.
31
A basic idea of grounded theory is to choose research participants who have lived through the phenomenon that you want to learn about. Having lived through the phenomenon, they are experts on it. Your research concern is to learn about their lived experience. – Auerbach & Silverstein, 2003, p. 15
4
A View of the Landscape In this chapter I present my Phase 1 study, which was a semi-structured interview study with parents and teachers of students with cognitive disabilities. The purpose of this study was to gain a broad understanding of the role that assistive technology (AT) plays today among these populations, the barriers to AT adoption and use, and the hopes and desires that participants have towards AT. I begin by discussing the high abandonment rate of AT, and I identify shortcomings in existing literature about AT abandonment and adoption for individuals with cognitive disabilities. Next I present the method and participants of the semi-structured interviews. I then present my findings, which are organized into three themes: understanding the adoption process; issues in current AT use; and hopes and dreams about future technology. In each section, I discuss design implications from the findings. I conclude with a discussion of the broader implications of this work in HCI, including the need for designers to consider the entire life cycle of the technology, from configuration, to maintenance, to upgrade and replacement. This research was published in [Dawe, 2006].
32
4.1. Introduction One of the first statistics one learns when introduced to assistive technology development is the surprisingly high percentage of AT devices that are purchased and later abandoned – 35% or more [Kintsch & DePaula, 2002]. One weakness in the existing literature is that studies of technology adoption tend to group together all types of disabilities, from motor and sensory to cognitive (such as [Martin, 1999; Reimer-Reiss, 2000]). The types of AT in use for different disabilities vary widely. A single abandonment rate for all AT does not provide a very useful picture for anyone given the large differences among the devices and the populations that use them. Another shortcoming with studies in this area is that analyses tend to focus purely on successful adoption versus abandonment. This simplistic, binary approach fails to reveal the ways and contexts in which technology is used [Suchman, 1987], and whether the technology is effectively addressing the impairment. This approach also fails to reflect the processual nature of technology adoption: the adoption process has multiple stages that take place over time [Rogers, 1995]. Previous research narrowly defines and identifies assistive technology. I argue that the category of assistive technology is problematic because it ignores “regular” technology that has been re-purposed or appropriated [Carroll, 2004] to compensate for a disability. By starting out with a preconception of what is and is not AT, these studies are only providing a partial picture of the role technology is playing in the lives of people with disabilities. As a researcher with little experience in AT and its users, I began my research inquiry with a broad interview study to gain some basic understanding in these issues. I conducted semi-structured interviews with parents and teachers of young people with cognitive disabilities. My core research questions were: •
What role does technology play today in the lives of families who have a child with cognitive disabilities?
•
How do families find, acquire, and use these technologies?
•
What are the key factors that increase or decrease adoption of technology?
33
•
What are the future visions – hopes and dreams – of these families regarding assistive technology?
This chapter begins with a description of the methods and setting of this study. I then discuss three major themes that emerged in the findings: the role of multiple individuals in the adoption process; the critical importance of simplicity of AT, particularly in configuration and support; and the perceived potential benefits of AT. I provide design implications for each of these issues and conclude with a discussion of the broader impacts of this research in identifying direction for further AT adoption and usage research. 4.2. Methods and Setting The goal of this formative study was to gain a broad understanding of the types of assistive technologies used today by young individuals with cognitive disabilities, the purpose and usage contexts of these technologies, and how these technologies are adopted. Since young people spend half their day in the school system, it was important to include both teachers and parents as interview respondents. Where possible, I interviewed the teacher and parent of the same child. Appendix A includes the interview guide. Interview questions followed five themes: 1. Current AT devices in use and their role and impact in the life of the student/child 2. Assistive technology devices that the respondents used in the past and why the devices are no longer used 3. Other technology (e.g., PCs, Walkman devices, video games) used by the student or child 4. Wishes for future assistive technology 5. Initial response to and recommendations for two Cognitive Levers (CLever) projects (MAPS, a handheld prompting system, and Mobility for All, a public transportation guidance system). The interviews were deliberately conversational in nature rather than formal surveys to let themes around these topics emerge naturally. Interviews were
34
conducted in the family homes or the teachers’ classrooms. This setting allowed the respondents to show and demonstrate many of the devices described, as well as the physical contexts in which they are used. As with many studies of minority populations, there are only a small number of potential participants in the local area. As such, the participant sample was nonrandom and identified through snowball sampling, where current participants identify other potential participants. The inclusion criteria were deliberately functional rather than clinical and describe an upper and lower bound of ability. Specifically, participants were chosen who have a student or child with the following characteristics. The student or child: •
Has moderate to severe developmental/cognitive disabilities
•
Is physically capable of independently operating very simple technology (e.g., a switch, which is an input device consisting of one large button)
•
Is pre-adolescent or older and is socially capable of performing tasks away from a caregiver
•
May or may not have experience using an AT device
Functional descriptions have been found to be more effective than clinical or condition-based descriptions to indicate whether an individual can use and benefit from AT [Newell, Carmichael et al., 2002]. Using simple, functional criteria provides caregivers the opportunity to make their own assessment about whether their child or student fit the criteria, and does not require access to clinical data in medical records. In total, 20 interviews were conducted with twelve families and eight teachers in the Boulder and Denver area. Interviews lasted 30 to 90 minutes. The family interviews were somewhat longer and teacher interviews shorter due to teachers’ constrained schedules. Among the families, I interviewed four mothers (the rest of the family was not at home) and eight mother-father pairs. The child with a cognitive disability was present for eight of the family interviews. All of the families were broadly middle class, held a variety of occupations, and their children attended public school.
35
The individuals with cognitive disabilities ranged in type of disability, age, and gender. Four of the individuals have autism, five have Down Syndrome, and twelve have other or unspecified developmental disabilities. Many of these individuals have multiple handicaps, meaning they have a physical, visual, or other disability in addition to cognitive impairment. The ages of the children ranged from 13 to 23, with the majority (13) between 18 – 23 years old. All children lived at home with their parents at the time of the interviews (since the study, two children have moved out and are now living with assistance in their own apartments). A variety of technologies were discussed, ranging from communication aids to alternate computer input devices to video games and Walkman devices. I summarize the different types of assistive technologies that were either tried or are currently in use by the participants in Table 4.1. This table does not include technology used purely for recreation (such as Game Boys and Walkman music players). Number of Devices Category
Example(s)
Communication
Used at One Time
Still in Use
Considered but not Used
Augmentative communication devices, Picture symbols
21
6
1
Writing
Word prediction & spelling software (e.g., AlphaSmart)
10
5
1
Prompting/ Scheduling
Picture schedulers, timers, watches
10
5
Reading
Screen reading software
9
5
Educational Software
On-line & software games
8
7
Alternative Input
Programmable Keyboards (e.g., IntelliKeys), switch
6
4
Math
Large button calculator
3
3
Reading & Writing
Write Out Loud
3
3
Entertainment
Talker used as a toy
1
1
Medical Monitoring
LifeLine System
1
1
Writing / Navigating Web
Voyager Suite by AbleLink Tech.
1
1
77
42
Total:
Table 4.1: Assistive technology devices discussed in interviews 36
2
Interviews were audio-taped and transcribed, and then analyzed qualitatively using methods from grounded theory [Strauss & Corbin, 1998]. Grounded theory is a structured approach to qualitative data analysis that allows concepts to “emerge” from the data (in this case, interview transcriptions), rather than an analysis based on a predetermined set of categories. For this study, I created codes for concepts observed in the data, which were iteratively reviewed, revised, and similar concepts combined, until a few key “theories” emerged. Throughout this process I sought to stay “close” to the data by continuously revisiting the transcriptions. Key themes emerged around the complexity of the adoption process, the need for simplicity in technology configuration and maintenance, and a common hopeful optimism towards technology.
4.3. Understanding the Adoption Process As mentioned earlier, studies have shown that there is a high rate of unsuccessful AT adoption. Through these interviews it became apparent that technology adoption is a process with multiple stages and cannot be understood by only looking at a single point in time. It became clear that breakdowns in adoption can occur due to the conflicting perspectives of the many individuals involved, and the length of the adoption process. 4.3.1. Role of Different Individuals at Adoption Stages Rogers [1995] presents an adoption process model with five stages: knowledge or awareness of the innovation, persuasion (internal or external) to adopt, deciding to adopt, implementing the innovation by incorporating it into one’s situated context, and confirming that the innovation is appropriate or not. The earlier stages are based on the intended or predicted uses of the technology, while the incorporation and confirming stages are based on the actual needs of the users in the context of their existing practices. However, Rogers’ model and many adoption studies assume that a single individual, or the same group of individuals, are making the decisions at each adoption stage. Among the interview participants this was often not the case, but
37
rather different individuals or groups were involved in the adoption process at different stages. One illustrative example is Erin (all names used in this dissertation are pseudonyms), who is a high school student with moderate to severe cognitive disabilities and with moderate visual and hearing impairments. Erin’s mother Paula felt that a touchscreen would provide the best interface to the computer for her daughter. However, the family did not have a touchscreen at home, so the daughter used the mouse which was more difficult for her. Meanwhile, Erin’s teacher stopped using a touchscreen with Erin because the teacher was aware that Erin had become accustomed to using a mouse at home; the teacher reported that Erin reached for the mouse instead of the touchscreen. In this case the teacher felt it was important to give the daughter a consistent experience at school and home; the parent wanted a touchscreen to be used at school because it was easier for her daughter. At first glance it may seem like the teacher and parents simply were not communicating with each other. However this example illustrates a common theme in the interviews: parents wanted the best technology available for their child and wanted it to be introduced and supported by the school system, which they felt had the expertise and support to do so. Busy teachers, on the other hand, gave a preference to technology that a student was already familiar with either from home or from previous school years. Teachers are also encouraged to give their students the most available and least costly technology options, given that there is a limited resource pool of AT in the school district. This example demonstrates how the priorities of the parents and teachers that affect technology choice in the decision-making stage may be in conflict with one another. In other examples, such as the case described below, teachers and parents played prominent roles at different stages of the adoption process, which created unintended challenges to adoption. 4.3.2. Length of Adoption Process For a number of families the acquisition of AT is a lengthy process, in some part due to working with external agencies such as insurance companies and Medicaid. In
38
the case of Kate and Nick, the acquisition of an AT device took over a year, during which time the potential user graduated high school and developed a lifestyle where the device no longer added much value. Nick’s mother, Kate, had the school system’s support while locating and choosing AT for her son, but lost support at the incorporation stage. Kate wanted a portable device that would help her son Nick, who has autism and is almost completely nonverbal, to communicate. Since Nick recently turned 18, Medicaid would pay for a device. Kate described how the school and the family worked together to identify a high functionality, expensive communication device for Nick called a LightWriter. Yeah those guys [the AT specialists], they came out. And we looked at all sorts of things together, and we tried all sorts of things, and it was the whole team. And Nick was there too ... and he tried them all too. And [the LightWriter] was ... the most portable, it had the most functions. People didn’t have to stand looking over his shoulder to read what he was saying, so that, it had that comfort factor, too. But things broke down in the incorporation stage when Kate thought she would have the support of the school system but did not. Acquiring the device through Medicaid took over a year, and by that time Nick was close to finishing school. Kate explained: When it finally came there wasn’t much time left in his school career to actually make it be functional. Because we all agreed that he needed to use it at school, in the school setting, to make sure that it wasn’t a toy. Because if he came home, you know how can you force it, at home? When you don’t really have to talk at home anyways. You know what I mean? So we all agreed that it had to happen at school. That school had to be the forcing agent here to use it, to make sure that it wasn’t a toy. And there just wasn’t time. Kate went on to describe how she feels stuck with what she has, and indicates that she questioned the appropriateness of the device in the first place: And once it was bought it was pretty much, this is it. It’s yours now. And you’re never getting another one. I kept saying, there has to be something more portable, there has to be something more portable. And I wasn’t shown anything. From all the catalogs that they had ... I don’t know how different it would have been if the timing had been better. It might have turned out to be really, more productive, if we could have actually had, like the school really working on it. As opposed to not. So, I don’t know. Who knows.
39
Since acquiring the $3,500 (USD) LightWriter, it has sat in Kate’s closet while Nick adopted a small and portal “hip talker” that Kate found on her own and cost approximately $200. Schools frequently work with families to acquire a device near the end of the child’s school career because when students turn 18 they have additional financial support for technology through Medicaid. However, this can create a situation where technology is chosen that is too complex and expensive than families would have otherwise chosen. Furthermore, families are relying on the school for help incorporating the device into the student’s life, but there may not be sufficient time for this. 4.3.3. Design Implications Among the interview respondents it was common for multiple groups to be involved in the adoption process, and play a different role at various stages of the process. Specifically, AT specialists or speech therapists often identified a potential technology; teachers were often involved in the trial period and decision-making stage; and parents were expected to incorporate the device into the student’s life (or maintain the AT after the student leaves the school system). Each of these caregivers has a unique set of motives and beliefs about the purpose and appropriate usage of technology and about a student’s abilities, needs, and interests. This can create conflicts in the AT adoption process. These findings are consistent with findings by Orlikowski & Gash [1994] and others who have demonstrated that when there are conflicting perspectives or framings of the situation by different stakeholders – in this case, between AT specialists, teachers, and parents – it can lead to deep misunderstandings and unsuccessful adoption of technology. To further understand adoption in this context, designers should consider the different stages of the adoption process and the network of caregivers involved. When collecting requirements for new assistive technology, designers must make the effort to understand and gather input from the various individuals who will be involved in the adoption process at different stages, including the often overlooked stages of incorporation and maintenance. For example, a designer might analyze the current
40
practices and patterns of usage of existing technology in different physical contexts, such as while an individual is at her job or volunteer position, or involved in a community activity. In addition, the designer should interview the individual’s caregivers in those environments to understand the different caregivers’ perspectives on the individual and the technology. The design process should include bringing awareness to, and reconciling some of the divergent expectations, perceptions, and goals among the network of caregivers. 4.4. Keep it Small and Simple, Please! Parents and teachers identified three major desirable features for increasing simplicity of assistive technology: portability, simple yet evolving functionality, and ease of upgrade and replacement. 4.4.1. Portability Nearly every parent and teacher cited small size and portability as a benefit of a current tool they use, or as an improvement on a bulky device. Some families described the lack of portability as a reason for abandonment. One mother, Vivian, explained why her daughter Laura never moved her communication device from the kitchen table, and how this lack of portability ultimately contributed to abandonment: It was cumbersome, that’s a cumbersome thing to haul around for a kid, and then the motivation has to be there. If you want to communicate, it has to work. It has to coordinate with what you want, what you are able to do, and try to get that message out there. Well, she could just as easily go over to that cabinet and grab a cookie, as try to come over here, and try to press on this super hock [Laura’s communication device] cookie button and it would say, “I want a cookie.” And we tried to do different things, to make it even more motivational, but this is a kid who pretty much gets her needs met without language. This is a case, representative of other parents’ and teachers’ descriptions, in which a cumbersome, heavy AT device was not used because the user was in a mobile environment. This example illustrates the challenges parents face in motivating their children to use AT when the children can get their needs met in a more direct way, although it may be less socially appropriate. This was particularly an issue with augmentative communication devices such as the device Vivian wanted her daughter 41
to use. Vivian wanted Laura to learn to communicate with strangers in a socially acceptable way and viewed technology as a way to facilitate that; but Laura saw the device as unnecessary and an annoyance at home where she could communicate more directly through action. Vivian felt that motivation may have increased if the device was something Laura could have easily carried around with her. 4.4.2. Simple to Use and Able to Evolve with a Child’s Abilities While it is a platitude to say that something is better if it is “easier to use,” parents and teachers strongly emphasized ease of use as an important property of AT, referring specifically to the need for a very simple interface and a low learning curve. A few respondents explained that because of their child’s disability, if their child had initial difficulty using the technology they were very unlikely to ever go back to it. Parents suggested that the device could increase in complexity over time to grow with its user as long as it started out very easy. When asked to give the most important factor in successful use of AT for her son, one mother explained: I think it would have to have ease of use, it would have to be easy to use … I mean, at first obviously it would take him a while to learn it, but definitely I think the biggest factor has to be ease of use. She went on to explain that the technology functionality should grow with the child: ... the device would have to also have the capability of becoming more complex as they get more and more used to the computer so that they’re still advancing with the system, not something that’s outlived its use. It has to be able to be expanded upon. Parents’ perception of technology complexity included understanding its functionality, documentation, getting it set up, and recovering from failures. Complexity did not always refer to the device interface itself being too complex for the child to use. For example, Kate explained why the LightWriter never got used by describing three areas of complexity: configuration, user interface, and documentation. When explaining the configuration process, she described, “it’s too complex, too many steps and too much work to just say his name and address.”
42
Kate explained her impression of the technical documentation that came with the device, and why she never bothered calling technical support: The book was so confusing that I couldn’t even figure out which version of Light Writer I had. I mean, so it was like, and you know if you call somebody, they’re going to ask you a question, and you’re going to sound stupid because you don’t even know what machine you have sitting in front of you! … I mean, it took me a long time to figure out that I don’t have the one that does this, I have the one that just does this. So that’s why I can’t get it to do what I’ve been trying to get it to do. So, while there probably was technical support, I never used it because – it was just too hard. Finally, Kate explained why she was afraid of too many buttons on the complex user interface, and how this additional complexity made it harder to troubleshoot: [I need] something without as many buttons, because he wouldn’t know how to use all the buttons, and he’ll jam it, and I would have to come home and figure out how to fix it. Perceived complexity of configuration also affected AT adoption in the case of Carin and Dan, who recently purchased a few expensive devices for their daughter Linsey through Medicaid. Linsey is in her early-twenties and has moderate cognitive disabilities. All of the recently purchased expensive devices required initial configuration, which Carin had not gotten around to doing. The assistive devices that the family did use with Linsey were a simple memo recorder that recorded a single 30-second message, and a kitchen timer that prompted Linsey when to leave the house. Carin explained why the more expensive, complex devices had not been used: This is very complex. And I think the issue is, getting the time to set it up and work all the bugs out … And the thing about the Voyager [an AT software system for browsing the web], or any of these, is just sticking with it, getting really good at operating it. If we had a computer set up … I mean, this is on a laptop right now. And with time con [sic] ... space limitations, and pulling it out, and setting it up, and this and that right now, it just hasn’t been a priority in getting that mastered. And I’m really the only computer nerd in the house, so to speak. Carin contrasted this with the simple memo recorder that she used every day with Linsey to prepare her for that day’s activities:
43
And what I like about this … is the simplicity of it. The ease of use, the, in the hurried world that we all live, it’s just real, there. She pushes a button, she gets a message, she knows what the day will include. In Carin and Dan’s case, the time investment of figuring out how to set up the more expensive devices they picked out for Linsey was too high, given that the AT had not yet shown any value. Carin’s desire for technology that is “just there” and reliable was echoed by other parents who struggled with maintaining technology at home. Another mother explained how her son’s simple communication device would periodically “zero itself out,” and she would have to re-program everything in from scratch, each time having to call technical support because the programming process was so difficult and complex. Over time the communication device was abandoned because of this painful process, and her son instead used paper-based picture cards to communicate. An interesting observation that adds nuance, if not contradiction, to this description is that while simplicity was repeatedly emphasized, many of the children discussed in the interviews were rather technically savvy, and were skilled users of everyday technology, including video games, computers, Walkmen, VCRs and stereos. In fact, many parents explained that their children were drawn to technology because of its predictability and consistent behavior; some parents and teachers even felt they had to limit the time their child spent playing video games or at the computer. A teacher described a case where a student quickly learned how to reprogram his communication device and would intentionally erase its contents, requiring the teacher to reconfigure it all over again. The teacher lamented that the technology designers underestimated the technical ability of the users. 4.4.3. Simple to Upgrade and Replace Simplicity also emerged as an important feature when the technology was being upgraded or replaced. The importance of being able to update or replace a system arose more among families than teachers, probably because school systems have a better process in place for updating or replacing broken technology. Parents recognized that a fundamental property of technology is that it will break or “wear
44
out,” and will need to be replaced. A mother described this in the context of her daughter’s large button calculator: She has a large number calculator. So both the keypad is large, and the display is large, and it’s tilted so it cuts down on the glare. And we got that at Walgreens! They’re not hard to find. We’ve gotten them from Radio Shack too. And when they wear out, they wear out, but they’re not very expensive ... plus she has one for home and one for school. Another mother explained why she decided to buy an inexpensive communication device for her son: It was under 200 dollars. So when it breaks, it will be easy – it will be replaceable. It won’t be like a big, ‘Oh my goodness. Now what do we do?’ The same mother described how she found a simple memo device at Radio Shack that worked effectively for her son, and bought every single unit in the store (!): I know in middle school we had a cool little thing that we used. Again, I found a $69 talking memo thing, at Radio Shack, that was the coolest thing, we used it in middle school. But when I found it, it was already on clearance. So I bought up every one I could find. And when they broke, they just broke, nobody would fix them. So we used that, and that was in middle school a lot. The ability to replace technology relates to issues of cost. Cost of technology played a different role at school and at home. As mentioned earlier, at school high cost devices actually seem to be abandoned more often because of the “limited resource pool” of assistive technology in the school district. For example, the school district may have five communication devices, and must decide which five students will benefit the most from having the device. If a device is not making an impact for a student, the teacher has pressure to return the device, rather than invest time helping the student learn and use it, because the impact the device is having with the student must justify it not being available for other students. One parent explained it like this: There’s usually kind of a friendly inertia working against you in the sense that yeah, they want to be nice to you but they really don’t want to do anything special because they have a limited dollar pool. Teachers themselves described how they work within the paradigm of a limited resource pool: 45
I was the one who terminated the use of the Dynavox, because I felt like the district was tying up a $7,000 piece of equipment and it wasn’t fair to other people. So that was my ultimate decision. Could he use it? Probably, two or three years, but, I didn’t know if, weighing the need for it in the district, if that was fair. Another teacher described how she stopped using a touchscreen with a student, and how the cost of the device was a consideration: And that’s a pretty expensive piece of equipment, and rather than have it getting broken in the classroom, I just said, she’s not using this anyway, so let’s send it back. At home, although many parents have some AT costs covered through insurance or Medicaid, cost is still an issue largely because parents did not want to spend a lot of money on a device that would break or get lost, which most parents assumed would happen. One mother explained why she did not buy her autistic son a PDA, even though he would probably benefit from it: And you know, he is pretty reckless, we didn’t want to have a piece of $300 equipment that he was carrying around, either. So expense became an issue. 4.4.4. Design Implications The various aspects of simplicity described here: portability, ease-of-learning, ease of configuration, and replaceability span the hardware and software of a system. Portability is largely dependent on the hardware choice; ease-of-learning and ease of configuration are mainly dependent on the software design; and reliability and replaceability really depend on both hardware and software. On the hardware side, replaceability is tied to cost and availability. These observations suggest two important implications for AT design. First, devices that do not have direct usefulness out-of-the-box, before configuration or customization, are less likely to be adopted. Parents are generally busy people and few enjoy “tinkering” with technology. Parents are more willing to invest time if they perceive value and usefulness in technology. Therefore, technology that can demonstrate immediate functional value without requiring a major investment of time and effort in learning configuration procedures will have a better chance at successful adoption. 46
Caregivers’ desire for technology that is simple to use, yet able to grow with the child’s abilities suggests a shift from viewing simplicity as a static attribute to a quality that is relative to an individual’s abilities and needs. Rather than reducing the problem to the notion that “simpler is better,” designers should strive to support both a low threshold and high ceiling [Papert, 1980]. This implies a simple and easy-tolearn out-of-the-box experience that can increase in functionality and complexity as the users’ skills grow. Appropriate simplicity can be viewed as creating an optimal flow in the user experience [Csikszentmihalyi, 1990], which balances an individual’s skills and desire to be challenged as represented in Figure 4.1 (recreated from [Csikszentmihalyi, 1990]). Too much complexity produces anxiety, but not enough functionality results in a lack of usefulness and boredom. As users’ comfort level and skills increase with technology, the optimal flow shifts and the technology can gain additional functionality.
Figure 4.1: Optimal flow diagram [Csikszentmihalyi, 1990] Second, the importance of supporting configuration, updating, and replacement in the software means that the system should support backing-up, exporting, and restoring a system configuration, including any customizations. Usability must include not only the functionality and UI of the device during normal use, but more broadly encompass the documentation and support materials and the lifecycle of the technology from configuration to replacement. This issue has been raised elsewhere [Ketola, 2004] in a call to expand the lens of usability to look not only at initial use, 47
but also at the “end of life” of hardware or new versions of software, when the software needs to be moved or upgraded. 4.5. Hopes, Dreams, and Technology Parents and teachers described desired outcomes of assistive technology in broad terms, including increased independence and safety, and increased and more appropriate social interaction. These desired outcomes reflect the role of technology, beyond merely facilitating a task, in altering the emotional and social context in which it is used. 4.5.1. Increased Independence Both parents and teachers cited increased independence as a major benefit to using technology, and one of the potentials that technology holds. For example, Donna is a senior in high school who is multiply handicapped and has cognitive, visual and physical disabilities. Donna can walk with the assistance of ankle splints. Cognitively, she has difficulty remembering multiple steps of a task. She held a workstudy job at the high school cafeteria serving food. Her mother, Emma, created a lowtech memory aid to assist Donna at work. She took pictures of Donna at her job and created a poster-sized chart with the pictures and single word reminders of Donna’s different tasks on the job. Emma described the benefit of the chart: She was able to be more independent, and knew, I have a chart, I don’t have to ask … it’s an independence thing. Parents discussed the challenge of finding the right technology for their children that would increase independence. Bobby is 20 years old with autism, and lived at home with his parents. He loved computers and calculators, but he did not use any technology that addresses his disabilities. His mother Mary expressed the difficulty of finding the right kind of AT to bring Bobby more independence: We’ve never gotten a handle on what would fit. We know that he could certainly use some support in a lot of different areas and could be more independent, I think, with support, but we don’t really know what that is.
48
Parents often combined independence with the need for social interaction. In the following example, a father explained that technology should increase his daughter’s independence as well as community involvement. Annie lived at home with her parents Jim and Sandy. Annie is 23 years old and has Down Syndrome. Annie used more technology when she was in school, but after graduating she just used her mobile phone and occasionally a computer. She was active in the community, and held two part time jobs. When asked how to measure the success of technology, Jim explained: What’s the definition of success ... if it’s used, and if it keeps her on task, and enables her to be more independent, and out in the community, then it has value. Jim described the important role Annie’s mobile phone played in bringing her independence: Well, when she first started going to work, I mean, she’d call from the bus stop, she’d call from the bus itself, she’d call when she got to Middletown, she’d call when she got to work. I mean, it’s part of the process. Without the cell phone, she wouldn’t have a job. And now, I don’t know Sandy, does she call a couple times a week, maybe? Annie’s mother went on to connect the importance of the social connection, and how it increased her sense of safety: Yeah it’s just, I know I can contact her, find her you know, and she can call us, or she can call ... she knows a lot of numbers, she knows a lot of people she can call if she needs help. Emma similarly described the potential role a mobile phone could play in her daughter Donna’s independence: It’s almost like an intermediary to being independent. I’m not with you, but I can follow where you are and I can contact you immediately and you can contact me immediately. In this example Emma articulated how the mobile phone could potentially increase social connectedness as well as her perception of safety, because the mother and daughter would only be a phone call away. Unfortunately her daughter Donna did
49
not currently use a mobile phone because the buttons of a standard mobile phone are too small, so this potential was yet unrealized. 4.5.2. Increased Social Interaction and More Appropriate Social Behavior Most of the AT that families used played a role in increasing social interaction. Another important corollary to this is that parents did not just want increased social interaction, but also appropriate social interaction mediated through the technology. For example, Kate found or devised a few simple technology devices for her nonverbal son Nick, all of which were directed at increasing his social interaction. Nick used an inexpensive communication device that has 8 buttons, each of which caused a voice recording to play; Kate programmed “hungry,” “hello,” “goodbye,” and other single word messages that he could use “out in the community.” To increase Nick’s social interaction, Kate made sure that Nick wrote and sent an email every day using a small laptop that she purchased inexpensively. Kate also “appropriated” a talking photo album for Nick that held recorded messages, along with pictures, so that Nick could introduce himself to others. Parents described how the right technology could guide their children to behave in ways that are more socially appropriate, and become more socially accepted. Laura, a freshman in high school, sometimes used a computer at school and at home, and liked to visit musical web sites (with the help of her mom, Vivian) and look at photographs on the screen. Vivian explained how technology might help Laura have more socially appropriate behavior: Appropriateness instead of maybe going up and trying to hug you, which she would be prone to do. You know, which a lot of people with disabilities do. And we try to teach her, okay, what would be more appropriate is high five, or what’s up. I see technology as maybe a way to make a friend. You know, just in the limited sense, but to just say hi, what’s up, how are you. Vivian was hopeful that technology could play a potential role in helping her daughter overcome existing social barriers and become more socially accepted.
50
4.5.3. Design Implications Communication devices that not only support communication but also mediate it according to social norms are found in assistive technology as well as mainstream technology. One example is the Aware Chair [Moore, 2003], an assistive device used by people who are “locked in,” meaning that they have very little physical ability to communicate, yet have a lot of brain activity. The Aware Chair has a simple biosensor input, and a smart interface that recommends communication options according to the time of day, who is in the room, and history of interaction with conversation partners. In another example, a “persuasive” mobile phone called the KITphone (keep-in-touch) reminds phone owners periodically to call people in their contact list who they have not been in touch with recently [Golder, 2004]. Assistive technology designers should consider the role that technology will play in caregivers’ goals around independence, social connection, and safety. Research in mobile phone usage has identified different roles that the mobile phone plays among different age groups: for young people it is primary for social expressiveness and social interaction; for families it is primarily for coordination; and for the elderly it is adopted primarily for safety, and “remote caregiving” [Ling, 2004]. It is important to study what types of communication tasks are involved in caregiver-care recipient communication, and design technology that supports those tasks. As a result of these findings, I have focused Phase 2 and Phase 3 of my research on: •
Understanding how mobile communication plays a role in addressing the goals of parents and individuals with cognitive disabilities
•
Exploring what technology might look like to support these tasks
4.6. Summary This research study focused on understanding technology usage among individuals with cognitive disabilities, in order to provide a more accurate picture than what is given in studies that group together all people with disabilities. This research methodology allowed participants to define which technology they find to be assistive, rather than starting with a problematic pre-determined definition of AT. My approach led to the observation that much of what families are using is re-purposed or 51
appropriated technology that was not intended to be AT, but has been found to functionally assist with the impairment. These off-the-shelf solutions tend to be simple, inexpensive and replaced without hardship. Studying the adoption process of existing AT among real families has uncovered important design implications for future technology. These implications apply not only to AT but are relevant more broadly to technology design in general. These implications include: •
Understanding technology adoption as a process. Adoption has multiple stages including technology identification, adoption decision-making, and technology incorporation. In many cases multiple stakeholders will be involved in these stages who will influence how the technology is adopted and ultimately used.
•
Providing a simple out-of-the-box experience. Technologies that required a lot of initial configuration before they could be used were often still sitting on the shelf; whereas technologies that provided simple, immediate usefulness were prevalent.
•
Expanding design focus to technology lifecycle. Studying the adoption process has reinforced the claim that user-centered design must focus on the entire life cycle of the technology, from configuration, to use, to upgrade and replacement. As users change and become more skilled with the technology, it should be able to adapt and support new requirements.
•
Potential for technology for increased independence, and social interaction. Assistive technology designers should consider the role that technology will play in caregivers’ goals around independence, social connection, and safety. Remote communication technologies emerged as a key technology area to support these goals.
52
5 Understanding Remote Communication Today In this chapter I present the second phase of my research. This was a semistructured interview study with young adults with cognitive disabilities and their caregivers. The study focused on current methods of remote communication, and perceptions of independence and safety. I first present the method, setting, and participants of the study. I then present the research findings, organized into four themes: nature of remote communication, challenges with mobile phones, issues around social involvement, and perspectives on independence. This study laid the groundwork for the technology probe study in Phase 3 described in Chapters 6 and 7. 5.1. Introduction In Phase 1, reported in Chapter 4, I conducted a formative interview study that provided a requisite view of the landscape of assistive technology in the lives of young people with cognitive disabilities with their caregivers. Communication
53
emerged as a critical application for current AT, in various forms including face-toface, written, and remote communication. Caregivers described the potential for better remote communication to increase independence, safety, and social interaction in the community, all of which are high priorities. In Phase 2, I took an in-depth look at how remote communication takes place today between caregivers and care recipients, focusing in particular on the role of remote communication in tasks related to safety and increased independence. I analyzed the data qualitatively and identified common themes in remote communication methods, purposes, and unsupported needs. 5.2. Methods and Setting Phase 2 was a semi-structured interview study with young adults with cognitive disabilities and their parents. Unlike in Phase 1, in Phase 2 I directly involved individuals with cognitive disabilities in the interviews, which required that the interview guide be more open-ended to adapt to the abilities of each interview participant. The interviews were longer than the formative interviews, lasting between two and four hours, and were usually broken up into two separate meetings. In this phase I recruited a smaller number of families to do a more in-depth study. I chose to recruit families with a child between 19 and 25 because this is generally an important transition period when the child develops a more independent lifestyle, and the parents replace the school system as the primary caregiver [Dawe, Fischer et al., 2005]. As described in Chapter 2, until the age of 21, individuals with cognitive disabilities receive full-time care from the school system. After age 21, parents usually become primary caregivers and are newly responsible for providing full-time care for their children including providing transportation, coordinating schedules (for employment, recreation, and social activities), and monitoring health and safety. Remote communication is crucial to connect parents who are at work with children who are at home or in a community activity. In total, I conducted nine interviews with 5 participant families. Two of the families had participated in the Phase 1 study. The other three families were recruited using snowball sampling, where participants of Phase 1 identified other participants
54
who fit the inclusion criteria. Table 5.1 describes each of the young adult participants, summarizing their age, gender, cognitive disability, and present use of remote communication technologies. The abilities of each of these participants varied across functional, physical, and social dimensions, but were within the range specified in Phase 1 (ability to do coarse motor operations on a PDA, with the social and functional skills to perform tasks away from a caregiver). The families were all from the Boulder and Denver area. Two families were single-parent households. The socioeconomic status ranged from lower to upper middle class. The parents held a variety of occupations. In four of the five families, the primary adult caregivers held full-time jobs.
55
Methods of Remote Communication Name
Description of Disability
Care Recipient
Caregiver
Alex male, 24
Cerebral palsy with moderate cognitive disability Ö Remembers numbers, can read Ö Good language production Ö Excellent language comprehension Ö High desire for social engagement Ö Low risk of getting lost Ö Mobility impaired
•
Frequent use of mobile phone (uses single-digit speed dial numbers)
•
Calls Alex’s mobile phone
•
Occasional use of landline (memorizes phone numbers)
Jake male, 21
Moderate – severe developmental delay, possibly also autistic Ö Sometimes able to remember numbers, can read Ö Moderate/low language production Ö Good language comprehension Ö Low desire for social engagement Ö Moderate risk of getting lost
•
No mobile phone
•
•
Rare use of landline (reads phone numbers)
Leaves hand-written notes
Linsey female, 25
Moderate – severe developmental delay Ö Little number recognition, cannot read Ö Moderate language production Ö Good language comprehension Ö High desire for social engagement Ö Low risk of getting lost
•
Frequent use of mobile phone (pushes redial)
•
Calls Linsey’s mobile phone
•
•
Occasional use of landline (asks someone to dial for her)
Leaves a recorded memo on memo device
•
Leaves message on home answering machine
•
Sets kitchen timers for scheduled activities
Down syndrome, moderate cognitive disability Ö Remembers numbers, can read Ö Good language production Ö Excellent language comprehension Ö Moderate/low desire for social engagement Ö Low risk of getting lost
•
Frequent use of mobile phone (memorizes phone numbers)
•
Calls Margaret’s mobile phone
•
•
Leaves hand-written notes
Occasional use of landline (memorizes phone numbers)
•
Sets kitchen timers for scheduled activities
Autistic, moderate cognitive disability Ö Remembers numbers, can read Ö Excellent language production Ö Excellent language comprehension Ö Low desire for social engagement Ö Moderate risk of getting lost Ö Sometimes socially inappropriate behavior
•
No mobile phone
•
Calls home landline
•
Frequent use of landline (memorizes phone numbers)
•
Leaves message on home answering machine (although Mark usually ignores it)
Margaret female, 24
Mark male, 19
Table 5.1: Young adults who participated in Phase 2 interviews
56
Participant families were all in the process of adjusting to the upcoming or recently past transition of their children out of the school system. During the time of the study, two of the young adults were under 21 and still in the school transition programs. All young adults were involved in life skills development programs including job training and recreational programs. Two of the young adults were in the process of moving out of the family home into their own apartments. One daughter had completely moved into an apartment, and had a live-in “companion” who provided part-time care in exchange for a reduction in rent. In the semi-structured interviews I covered a list of questions that went into depth about current remote communication, safety, and independence. I also explored if and how parents monitored their child’s activities in ways other than direct communication. Appendix B includes the interview guide. As in Phase 1, I did not attempt to ask the questions in the same order, but rather let the interviews flow like a conversation. I let the discussion move dynamically based on the respondents’ answers and interests, recognizing that unanticipated dimensions about these themes may emerge. Each interview session lasted from one and a half to two hours. The individuals with cognitive disabilities were not present for the entire interview because of the time duration. I began the interview by talking only with the parent(s). There was a second, shorter phase when the daughter or son participated in the presence of the parents. When talking with the individuals with cognitive disabilities, I discussed the same themes: remote communication, independence, and safety. I strived to keep the questions very simple and clear. The interview data was analyzed qualitatively, once again using grounded theory [Strauss & Corbin, 1998] to allow the themes to emerge from the data. I have organized the findings into the following general themes: nature of remote communication, challenges with mobile phones, desires for social involvement, and perspectives on independence.
57
5.3. Nature of Remote Communication In this section I discuss the ways that caregivers and care recipients accomplish remote communicate today and the various technologies families use to support this task. I then discuss who is involved and the reasons for remote communication. 5.3.1. Current Methods of Remote Communication Three of the five participant families were currently using mobile phones to communicate between caregiver and care recipient. The three young adults who were using mobile phones were able to receive calls and make calls in a limited way. The families with mobiles described a “cell phone lifestyle” in which the mobile phone was a central communication media between all members of the family. One family’s latest Christmas card was a family picture with each family member smiling and holding a mobile phone to their ear. Parents described how mobile phones were the primary way they communicated with and kept track of all their children. Other ways that caregivers and care recipients communicated with one another remotely included: calling on regular phones (“landlines”), leaving answering machine messages, leaving handwritten notes, leaving memos recorded on a stationary memo recorder, and setting alarm clocks and kitchen timers to indicate when the care recipient should leave the house for an activity. As in Phase 1, I observed that caregivers use many low-tech, simple devices in addition to mobile phones to keep in touch. Four out of five individuals with cognitive disabilities initiated remote communication on a regular basis. They initiated communication through mobile phones and landline phones. All of these individuals understood voicemail and frequently left voicemails for their parents, although none were able to check their own. 5.3.2. Who is Contacted and Why The individuals with cognitive disabilities who initiated remote communication mainly called their parents. Two individuals who were more socially oriented called
58
recipients included extended family and a few friends, although this was much less frequent than calls to parents. The purposes of the calls included many of the same reasons all families use mobile phones [Ling, 2004; Palen & Hughes, 2006]: reminding of upcoming events, scheduling, and coordination. Parents explained that calls frequently took place before, during, or after activities. 5.3.2.1. Safety Check-Ins and Locating Each other One reason these parents called their children – which is probably much less common among typical parents with their grown children – was to make a safety check-in. I define a safety check-in as a call whose primary purpose is to confirm the safety and well-being of the call recipient, rather than to exchange information. For example, parents explained that a frequent purpose of a call was to locate their child and make sure they are where they are supposed to be. Another example of a safety check-in is when parents are away from home and call their children who are home alone. Alex’s mother Deborah explained that she makes a brief call to Alex about once an hour while he’s home alone, “just to see that he’s fine.” Similarly, Mark’s father explains how he calls his son every day while he is at work to make sure that Mark arrived home safely from his community activities: Every day he’ll call and I’ll call him every day. Make sure he made it in where he was supposed to be, at the time he was supposed to be there, you know that kind of thing. I hate having a lost kid, especially one who might not get a hold of you if he does get lost. So I check up and make sure that he’s where he’s supposed to be. Among the two families whose children did not carry a mobile phone, the ability to make safety check-ins was the primary desired purpose of a mobile phone and these parents said they really missed having this ability. They cited concerns of losing their child as one of the main motivators for a safety check-in. Jake’s mother described the emotional impact of not being able to check in on Jake when he’s home alone, because he does not respond to a normal phone ring: As far as talking to him and seeing that he’s okay, it’s a real problem because, I learned this year, I’ve come home twice to asthma attacks ... and I had no way of 59
knowing, because he would not call, and tell me, mommy, I’m having an attack … It is frightening. The mental stress I go through not being able to communicate, or say, ‘Hi honey, I love you, how are you doing today? I just wanted to touch base and see if you are okay,’ you know, that’s missing.
5.3.2.2. Getting Un-Stuck: Calling when Plans Change One of the main reasons young adults with cognitive disabilities called their parents was to ask for help when something unpredictable happened in a plan, such as missing a bus. Parents described their children’s reduced ability to compensate for a changed plan, and how they would quickly “get stuck” in an unanticipated situation. Deborah explained how Alex has a limited ability to handle unexpected situations when they come up: he has maybe one or two back-up plans to handle an unforeseen situation. One of his parents’ big concerns is that the back-up plans will not work out, Alex will “get stuck,” and will not know what to do next. Mark’s father explained how Mark, who does not carry a mobile phone, has trouble with the unpredictability of the bus, and how he could easily become confused or potentially lost: He doesn’t go outside his normal route. He says, I know how to get from here to there, and back again, I don’t need anything else to help me do that. I already know the way. Vary that any, like ‘Oh can you go to the bank and make a deposit for me please?’ You know, something like that, um, no. That would be Lostville. That would be, I am not in familiar territory, therefore I am lost. But it would be nice to have him more independent so he can go more places as need be. Right now he is bus trained to certain locations only. And he usually doesn’t take a bus back from aikido; he walks to Mother’s rather than taking the bus back. Expressing his frustration with the unpredictable bus schedule, Mark responded in defense of choosing to walk rather than waiting for the bus: “Because that way I’ll actually get there at some point. I’ve waited for that bus before.” Similarly, Margaret’s parents explained that she calls them from the bus stop whenever she thinks she has missed the bus. Margaret’s mother described the last call she received from Margaret:
60
She was waiting for the bus on South Main road, and she said it was late, but I think she missed it. No, maybe it was late. I think they changed the schedule. So it was running about 10 minutes later than she thought. So she panicked, you know, and called me. Parents whose children carried mobile phones relied on their children’s ability to call if something came up or if the situation changed from the plan. Parents whose children did not carry mobile phones developed other strategies for checking in on their children, such as calling landline phones at certain times then the children were expected to be there. Yet they wished their children had the ability to call them if things did not go according to plan. 5.3.3. Role of the Caregiver Network A recurring theme from the Phase 1 study was the prominence of a caregiver network surrounding the young adult with cognitive disabilities. I found that individuals with cognitive disabilities primarily communicated remotely with their parents and then parents coordinated activities among the rest of the caregiver network. When the care recipients left the school system at 21 years old, parents replaced the schools as the nucleus of the caregiver network. Parents coordinated with other care providers including: case managers, job coaches, places of employment, personal assistants (who help with tasks such as laundry, grocery shopping, and bus training), friends and other family members, bus drivers, and coordinators of community program (such as Special Olympics). Parents were very in tune with their children and generally memorized their daily schedules. Parents communicated frequently with other caregivers to schedule and coordinate activities. In addition, parents arranged social activities with their children and children’s friends. 5.4. Challenges with Mobile Phones In this section I discuss challenges with off-the-shelf mobile phones. I describe the reasons why two of the young adults did not use a mobile phone. For the three mobile phone users, I outline the challenges they experienced with their phones, and how they used them in a fairly simple, limited way.
61
5.4.1. Reasons for Non-Use Two young adults in this study, Jake and Mark, did not use mobile phones. Their reasons for non-use were different but overlapping. Jake neither made calls nor answered them. His mother explained that Jake did not connect the ringing of the landline phone with his mother trying to contact him. She also thought he might have some anxiety about not knowing who is calling. Although they have a very close relationship, Jake never called his mother. She explained that he did not initiate calls because he did not connect the phone with the ability to reach his mother when she was away. His mother explained: I have tried for so many years to get that kid to communicate and to get him to talk to me when I’m away, you know. We’ve tried everything I can think of, and nothing seems to work, because he’ll just turn his back. He’s not interested in it. And you can get a special needs child to do miracles if they’re interested, and nothing if they’re not. Jake very rarely used phones at all. He had recently begun to use the landline for a very specific purpose: when he missed his class in the transition program, he was obligated to call the school to report his absence. He rehearsed making this call with his teachers. With extensive training, he felt obligated and confident enough to make the call on his own while his mother was at work. In contrast, Mark regularly used a landline phone and was capable of using a mobile phone. His reasons for not using a mobile phone were a combination of cost, phone reception, and a lack of motivation on Mark’s part to use it. Mark’s father, a single parent, explained that the family was on a tight budget and because he received no assistance from the government to pay for a mobile phone for Mark, he could not afford it (no one in the family had a mobile phone). Another factor was phone reception; the family lived in a rural area in the mountains and did not get cellular reception at home. This reduced the utility of the phone, although Mark spent much of his day in the city of Boulder where he did have reception. Mark also demonstrated a lack of motivation and even some concern about using a mobile phone. When I asked if he would like to have a mobile phone, he vaguely responded, “No, I’ve heard bad things about those cell phones.” Mark’s father
62
explained that he generally will not check answering messages at home because he assumes that if it is important, his father will call again later. 5.4.2. Difficulties for Mobile Phone Users The young adults who did use a mobile phone faced a number of software and hardware-based challenges in use. Parents emphasized the need for simplicity, echoing interview respondents in Phase 1. The young adults with cognitive disabilities did not use any of the features of the phone beyond making and receiving calls in a rudimentary way. Linsey could only receive calls and use the redial button to call back the last caller (she was not able to use the phone menus and did not have adequate number recognition to dial a number). Margaret was able to receive calls and make calls by memorizing a handful of numbers. Alex received calls and dialed using single-digit speed dial codes that his mother had set up for him (e.g., he remembered that his sister was “4”). His mother had set up a contact list with pictures and names, but Alex rarely used this menu because it was more difficult to access. In addition to overly complex mobile phone menus, these individuals struggled with plugging the phone into the charger and otherwise caring for it. All of the parents explained that they had to seek out a phone that did not have a delicate charger input because their children had difficulty with the fine manual dexterity required to plug in the charger. In general, they usually gave their children older mobile phones that were larger and more rugged, with bigger buttons and fewer options. Parents described how they made an effort to reinforce behavior to keep the phone functional and safe. Remembering to charge the phone was the most common example. Most parents created a ritual for their children to plug in their phones before going to bed. Parents developed strategies to keep the phone safe while the child was traveling by storing it in a special location in a backpack or purse, or mounting it in a safe place. Alex used a wheelchair that had a special location where he mounted his mobile phone.
63
5.5. Desire for and Challenges of Social Inclusion All parents in these interviews expressed the desire for their children to be more socially involved and have more friends. While the mobile phone played some role in this, it had limitations because the individuals with cognitive disabilities could only call a few people. Linsey was only able to call using the redial button, and thus she could only call the person who last called her (usually one of her parents). Margaret’s calls were limited to the phone numbers she could memorize. Alex was the only individual who used a notion of a contact list, and this was limited to 10 contacts since he used a single-digit speed dial number. Beyond the limitations of the mobile phone, parents described personal and social reasons why maintaining social connections was difficult. One reason was a lack of motivation and initiative on the part of the children. Some of the children truly enjoyed talking on the phone (including Linsey and Alex), while others treated it more as an obligation or necessity for getting their independence and as a lifeline when they needed something (as Margaret and Mark did). This seemed to be correlated to the children’s general level of interest in social interaction. Some of the children were fine staying home alone watching TV and movies or playing video games; others quickly became lonely and craved social contact. Parents encouraged their children to be more social. For example, Alex’s mom Deborah explained that Alex had a long-term girlfriend, Annie, who was one of the speed-dial numbers in his phone. Yet unlike many young men who drive their parents crazy by talking on the phone too much with their girlfriends, Alex would regularly forget to call his girlfriend. Deborah kept track of when Alex called Annie so she could remind him to call her every few days if they fell out of contact. Mark’s father explained how Mark did not have any interest in using the phone to maintain social connections. When asked if Mark made social phone calls, Mark’s father responded: No, social calls no. I can guarantee you zero social calls. Because he is not social. Does he have any friends? No, not one. Does he want to make friends? No, not one. Is he happy about the fact that he doesn’t have any friends? Yes.
64
Margaret’s mother similarly characterized her phone usage as functional, rather than social, but suggested (in a hopeful way) that someday Margaret may use the phone in a more social way: She doesn’t chat ... it’s not like the friends are calling back and forth. And I guess she could call her uncles, and stuff like that, but, she’s happy with it, you know, right now. When she gets older, she might want to keep in touch, like with her brother and sister when they’re gone. Another challenge to the children’s social involvement is the dedication required by other families in creating and maintaining a “circle of friends” among their children. Linsey’s mother Carin explained that even though Linsey was very social and loved to be involved in social activities, she was concerned that by moving into her own apartment Linsey might lose some social opportunities. She explained: [Linsey]’s a social person. There are times when I think it would just kill me, Melissa, to see her do nothing but that all day every day [motioning to Linsey who is watching TV] … I think I have done a good job of not letting that happen. But … the circle of friends is always the big thing. Having that social connection. This is a wonderful place [referring to Linsey’s apartment] and a great opportunity, but you’ve got to be feeding that social, I mean keeping that alive. Being in some sort of a group home, or community living, whatever, might be a good place for her for building friendships more.
5.6. Perspectives on Independence One of the goals of this interview study was to develop a better understanding of the perspectives of parents and children on the children’s increasing independence as they grow older. This emerged as a major theme in the Phase 1 study. This theme was especially appropriate for the parents in the Phase 2 interviews, because their children were going through a major life transition out of the school system and into the family care model. I found that while parents viewed increased independence as an important goal for their children, they were not rushing to create the distance themselves. In general they viewed their child’s care as a lifelong commitment. I also learned that the young adults showed a desire for increased independence.
65
When asked about whether their children would be more independent in the next five years, parents generally responded with “yes, but I don’t know how.” Not only did they lack a specific plan, they had trouble envisioning ways in which their child would have increased independence. For example, Mark’s father had trouble imagining how Mark’s job would change in the future, and struggled describing his son’s future independence: I don’t know. I think he’ll, hopefully improve at his job, get more hours and stuff, but his job’s heavily modified at Super Foods [grocery store where Mark worked], and they don’t really need him more, I don’t know. I really don’t know. I have no answer. Parents commonly expressed a feeling of their children’s independence as something they know is important, but do not work on enough. For example, when I asked Margaret’s parents if they are working on anything related to Margaret’s independence, Margaret’s father replied, “you mean, what should we be working on.” All of the parents in these interviews had extremely close relationships with their children with cognitive disabilities, and viewed the care of their children as a lifelong commitment. When I asked Jake’s mother if she foresaw herself as Jake’s primary caregiver for quite a while, she responded: Oh yes, oh yeah. We believe in family being with family, not family putting another person in a home. So we try very hard to take care of our home. Linsey’s mother expressed her confidence that she and Linsey will remain close, but also indicated her desire to have more time for her own life: Oh I think I’ll always be connected. I know I will. But I hope it does change so I can live my life and do some things for me, because I’ve given her about 25 of my last 52 years. Like many US teenagers and young adults, some of the young adult participants demonstrated some level of “breaking away” from their parents and finding their own independence, while parents struggled balancing their child’s independence and safety. Not all participants exhibited this behavior. For example, Linsey resisted independence when it meant a reduced social connection.
66
Jake’s mother described how he refers to his room as his apartment, and she would often lose Jake in the world of his video games. She explained: He’s very independent. He has set up his own, he calls it an apartment, ‘This is my apartment, Mommy.’ And he has his computer, and all of his game systems and his TV and everything in there … On Friday night and Saturday night he will stay up all night long, and go to bed about five in the morning. He plays his video games. That’s his world. Those are his people. At the time of the interview, Margaret’s parents had recently found an apartment for her, located a few blocks from their house. They let her spend the day there, but were not yet comfortable with letting her stay the night. Margaret on the other hand was very excited and anxious to stay the night alone in her apartment. When I asked Margaret if she would someday like to live away from her mom and dad, she responded by saying: “Well, I love them. Sometimes [smiling at her parents]. But, I feel like I want to have some space in my apartment…that’s what I’m doing tonight and Thursday!” Margaret’s statement was actually an assertive request to her parents for permission to stay at her apartment. Her parents responded by explaining to Margaret that she needed to learn a few more safety rules before she could stay in her apartment alone. 5.7. Summary Phase 2 allowed me to understand “how things are” in the lives of families caring for a young adult with cognitive disabilities, and how remote communication plays a role in safety, independence, and social connection. I grouped the findings into the following themes: •
Nature of Remote Communication. Young adults with cognitive disabilities communicate remotely with their parents on a frequent basis. Common reasons for communication were safety check-ins and locating each other, as well as recovering from a change of plan.
•
Challenges with Mobile Phones. Reasons for non-use included inability to cognitively associate the sound of the phone ringing with a request for communication, fear of unknown callers, and lack of motivation. Difficulties using off-the-shelf mobile phones included the inability to navigate and use 67
contact lists, difficulty plugging in the charger, and the small size and fragility of handsets. •
Desire for and Challenges of Social Inclusion. Parents desired increased social interaction for their children. Barriers to this included the inability of young adults to use the contact lists on their phones, the amount of effort on the parents’ part to maintain a “circle of friends,” and lack of motivation on the part of the young adult with cognitive disabilities.
•
Perspectives on Independence. While parents wanted their child to be more independent, safety concerns and strong emotional bonds meant they were not rushing into distancing themselves from their child. Some of the young adults wanted more independence from their parents, while others resisted independence when it meant a reduced social connection.
This study informed Phase 3, which explores “how things can be” by introducing a remote communication system customized to the needs of each young adult with cognitive disabilities. Phase 2 was an important precursor to this work, in that it developed a rapport with the families who continued into Phase 3. It established a level of trust that was essential for the participatory design activities that took place in the privacy of the home. Most importantly, it enabled me to begin the design activities with a solid understanding of where the families were coming from, including their current practices, priorities, desires and needs.
68
Practitioners recognize that all methods impose perspectives on reality by the type of data that they collect, and each tends to reveal something slightly different about the same symbolic ‘reality.’ – Brewer, 2000, p. 76
6
Co-Designing a Remote Communication System 6.1. Introduction In this highly-intensive phase of the research study, I worked with two families to co-design a remote communication system that their daughters used for 8 - 12 weeks. I begin by describing the research goals of the technology probe study, and how the probe method was adapted for the unique aspects of this design environment. Next I describe the activities and data collected before, during, and after the technology probe use period. This chapter provides details of the design method of the technology probe study. Chapters 9 and 10 present analyses of the data collected through this study. 6.2. Research Goals in Phase 3 The research goals for the technology probe study were: •
Ethnographic goal: to gain a deeper understanding of the nature of remote communication between young adults with cognitive disabilities and their 69
family and other caregivers in the context of supporting independence, safety, and social connectedness; and to understand if and how communication and support changes when new technology is introduced •
Engineering goal: to study the use of a touchscreen PDA platform in a realworld, mobile setting by individuals with cognitive disabilities
•
Design goal: to create a remote communication system that is used successfully by the participants, reflects their evolving needs and desires, and informs the requirements for a meta-design environment
6.3. Adapting the Technology Probe Method I adapted the traditional technology probe method described in [Hutchinson, 2003] to account for the unique aspects of the design environment surrounding individuals with cognitive disabilities. These design considerations were based on my findings in Phase 1 and Phase 2. Specifically, three considerations motivated adaptations to the method: •
Unique abilities of each user. A designer can make far fewer assumptions about the standard abilities of their users, and instead must rely on the information provided by caregivers and individuals with cognitive disabilities themselves.
•
Importance of simplicity. Technology that is too difficult or complex to use will not be adopted and can create a bad experience that can be a barrier to future technology adoption. As the user gains experience with technology, more functionality may be added to support new abilities.
•
Changes in user and usage environment. The environment in which the communication system will be used is more dynamic than a typical technical environment, and changes in users’ abilities and environment will introduce new and changing requirements on the technology.
To account for (1) and (2), I combined participatory design with the technology probe method and engaged each of the families as co-designers of their own technology probe. Traditionally, the probe is not viewed as a design activity in itself
70
that would involve end-users. In my environment, I needed to ensure that the probe would be usable and appropriate to the abilities of each user. I needed to rely on families as the knowledge experts. There was also an element of participant motivation influencing the design. This study required a high commitment level for busy families who do not have time to tinker with technology. The probe needed to be more than just fun technology: it needed to serve a useful purpose in the eyes of both the care recipients and the caregivers. To account for (3), I adapted the method to include the process of modifying the probe during use. While the traditional probe method recognizes that the technology will change the families’ patterns of communication and their perception of technology over time, the traditional method does not include modifying the probe during the course of the study. In fact, Hutchinson et al. [2003, p. 19] assert: Technology probes are not primarily about usability in the HCI sense. They are not changed during the use period based on user feedback. In fact, a deliberate lack of certain functionality might be chosen in an effort to provoke the users. For prototypes, usability is a primary concern and the design is expected to change during the use period to accommodate input from users. The authors were not primarily interested in how technology must evolve during use-time to support users’ changing needs. In contrast, understanding evolving functionality and customization needs was a central goal in my study. Creating a technology probe that evolves during the use period does not significantly change the spirit of the probe methodology, which is to inspire users and inform the design of future products. To inform the process of adapting a technology probe to act as an evolving artifact, I draw from the seed – evolutionary growth – reseed (SER) framework [Fischer & Ostwald, 2002] from the meta-design approach. The technology probe can be viewed as a “seed” in the SER model. The initial seed had simple functionality, grounded in remote communication needs identified in Phase 2. As families begin to incorporate the new technology into their lives, they identify ways that the device does not fit their needs, or ways that the device could be appropriated for other purposes. The families may change their practices as a result of the probe and develop new or different needs. This represents the “evolutionary growth” stage of the model. 71
The device is then “re-seeded” through active participatory design, and changes are implemented by the researcher. Figure 6.1 illustrates the technology probe as a seed in the SER evolution model.
Figure 6.1: Seed – evolutionary growth – reseed (SER) model applied to the remote communication probe study While the functionality of the probe is robust and tested for real-world use, the features are intentionally underdesigned in order to act as seeds for ongoing evolution. Underdesign in this context means that: •
Features are simple and open. The features are kept to a minimum and carefully chosen so that the probe is easy to use. The feature functionality exposed in the UI is robust and well-tested, but there may be some aspects of the functionality that are not yet implemented, such as configuration and data management.
•
Change is incremental. When a new idea emerges, we introduce it in a very simple way with only the core functionality needed to support it.
6.4. Method and Setting This study used a combination of research and design methods in a case study approach to understand the contextual, nuanced needs of individuals with cognitive disabilities (and their caregivers) when using remote communication technology. The participants were two families in the Boulder area who have young adult daughters 72
with cognitive disabilities. These two families had participated in Phase 1 and Phase 2 of the study. In Phase 3, each family co-designed and used a remote communication system for an extended period of time (8 – 12 weeks). I describe the study activities in detail below, in terms of three time periods: before, during, and after the technology probe use period. 6.4.1. Probe Functionality The technology probe was a PDA-based remote communication system, used by the individual with cognitive disabilities. The initial probe was designed to have a very simple interface and specifically support the most important remote communication tasks identified in Phase 2. While the specific design and functionality of the probe was determined separately for each family, I began Phase 3 with a good idea about what the probe should support based on the findings in Phases 1 and 2. The probe implementation supported the following activities: •
The user calls one or more primary caregivers using a picture-based menu
•
The user receives a call from one or more primary caregivers, and knows who is calling through a picture display (e.g., picture of the caller) and audio output
•
The device displays the time in a format appropriate for the user
The software implementation of the technology probe was a modified version of software by AbleLink Technologies [AbleLink Technologies, 2007], an assistive technology company in Colorado Springs, Colorado. At the time of the study, AbleLink had recently developed a prototype of a PDA-based mobile phone targeted specifically for individuals with cognitive disabilities. AbleLink also had software that supported scheduling and time-based reminders. Leveraging AbleLink’s code allowed me to begin with a stable, tested platform, on which I implemented UI and functionality modifications as well as other new features. The implementation of the technology probe is described in Chapter 7. Quantitative and qualitative data were combined in Phase 3 to understand how the participants used the probe. The data was iteratively reviewed and analyzed to inform the ongoing design changes. At the end of the study, the data was analyzed in more depth. More analysis can be found in Chapter 8. 73
6.4.2. Cost and Risk Considerations The participants received no monetary compensation. The remote communication software, hardware, and phone service (including unlimited minutes and a national calling plan) were provided free-of-charge to the families. At the end of the study, the families were given the hardware and had the option to continue using the software free of charge, supported by AbleLink Technologies. If the families chose to continue using the new phone, they received assistance transferring it over to their existing phone service plan. One risk in this study was that the new mobile phone could fail when the user was trying to initiate or receive a call. The new system could be less reliable or harder to use than the phone previously used by the young adult with cognitive disabilities. To minimize the risk of failure, users also carried the mobile phone that they used before the study, in case of emergencies. If the new phone failed to work or proved too hard to use, the user was able to fall back to his or her previous device. 6.5. Participant Families Both of the families who participated in the Phase 3 study had a daughter in her 20’s with moderate cognitive disabilities. In order to give the reader a realistic impression of each family and still protect the privacy of the participants, I created fictitious family names and used modified images to represent Linsey and Margaret. The participant families are described in Table 6.1. Primary Users
Parents
Additional Caregivers
Linsey Roberts female, 25
Carin, mother Dan, father
Abby, live-in companion
Margaret Smith female, 24
Sally, mother John, father
Donna, care companion
Table 6.1: Participants of the Phase 3 study
74
6.5.1. Linsey and the Roberts Family Linsey Roberts (Figure 6.2) is a 25 year old woman with developmental disabilities. She is the only child of Carin Roberts, 52, and Dan Roberts, 54. The Roberts are a middle-income family living in a suburban neighborhood near Boulder. Carin works full-time in IT at a large health care company. Dan Roberts works full-time in construction management, and often leaves early and gets home late because of long commutes. During the study, Linsey moved out of her parents’ home into her own apartment in the same neighborhood. Carin found a roommate for Linsey who provides part-time care in exchange for a reduction in rent. Abby, 28, moved into Linsey’s apartment as Linsey’s roommate and care
Figure 6.2: Linsey Roberts, talking on the phone to her father while on a vacation cruise with her mother Carin (photo blurred to protect privacy)
companion at the beginning of the technology probe study. Carin and Abby were the primary caregiver participants in the weekly interviews and participatory design activities. Table 6.2 summarizes Linsey’s functional abilities, focusing on the abilities that affect how she interacts with a mobile phone and how she participates in community activities.
75
Linsey Can:
Linsey has Difficulty:
Connect socially with people on the phone
Learning phone numbers, using mobile phone menus
Make lots of friends
Understanding social cues to get off the phone, not call people too much
Remember to charge her phone
Understanding why the mobile phone is dead when it has not been charged
Remember her daily schedule
Telling time
Hold a job (3), ride the bus independently
Reading, consistently recognizing numbers
Table 6.2: Functional description of Linsey's abilities Carin Roberts likes to keep herself and Linsey very busy. Carin frequently assumes a leadership role in work and community organization activities. Carin puts a lot of effort into keeping Linsey busy as well. Linsey has three part-time jobs: she rolls silverware for a restaurant, cares for children in a daycare facility at a local health club, and passes out water to residents of a retirement home. Carin signs Linsey up for many additional community activities including cooking and exercise classes, Special Olympics, and other social and recreational activities provided by a local organization that supports individuals with special needs. Linsey has received bus training as part of the government-funded services she receives. Linsey regularly takes the bus to the neighborhood recreation center to take exercise classes. Carin or other caregivers drive Linsey to most of her other activities. In her free time, Linsey enjoys talking on the phone, shopping, watching sports and weather on television, and spending time with family and friends.
76
6.5.2. Margaret and the Smith Family Margaret Smith, shown in Figure 6.3, is a 24 year old woman with Down syndrome. Margaret is the oldest of Sally and John Smith’s three children. The Smiths are a middle-income family who live in a suburb near Boulder. The Smiths and Roberts live in the same community and know each other. John Smith works full-time in marketing and is frequently out of town on business. Sally Smith is a full-time mother of three who manages the household. Sally’s other children (Margaret’s siblings) are Jane, a 22 year
Figure 6.3: Margaret Smith (photo blurred to protect privacy)
old graduate student living in Denver, and Tom, an 18 year old university student in Montana. At the beginning of the Phase 3 study Tom graduated high school and moved out of the family home. In the year preceding the Phase 3 study, the Smiths began renting an apartment near their home for Margaret. At first they sought to find a roommate/care companion for Margaret, similar to Linsey’s situation. Later they decided that finding and replacing roommates would be too much work in the long run and decided to teach Margaret to live alone. During the Phase 3 study, Margaret was transitioning to staying the night in her apartment, but was staying at the family home most nights. Table 6.3 summarizes Margaret’s functional abilities, focusing on the abilities that relate to her use of a mobile phone and her community activities.
77
Margaret Can:
Margaret has Difficulty:
Make phone calls to her parents by memorizing phone numbers
Using mobile phone menus
Enjoy spending time with people she likes
Maintaining contacts with friends and initiating social phone calls
Hold a job as a courtesy clerk at a grocery store
Receiving orders from her manager at the grocery story
Remember to charge her phone
Understanding why phone is dead when it has not been charged
Remember her daily schedule and tell time
Understanding longer time periods and knowing when to prepare for upcoming activities
Read and write
Understanding abstract and/or complex questions and conversation
Table 6.3: Functional description of Margaret's abilities Margaret holds a part-time job as a courtesy clerk at a grocery store. Although not as active as Linsey, Margaret participates once a week in a recreational outdoor program, and several times monthly in social activities through a local organization that supports individuals with special needs. In her free time, Margaret enjoys staying in her apartment alone, watching television and movies, and writing in her notebook. Margaret spends two afternoons a week with her care companion Donna, 61, who helps Margaret develop life skills such as grocery shopping and cleaning her apartment. 6.6. Design Activities before the Technology Probe Use Period Before the technology probe use period, the families engaged in a number of participatory design and assessment activities to design the remote communication technology probe. These activities iteratively generated more concrete artifacts representing the UI and functionality of the technology probe. The design activities intentionally began with very low fidelity artifacts that could be easily created, changed, and discarded as needed. This was to avoid biasing the participants and narrowing the design space by showing them an implementation or high-tech prototype too early. In each iteration, the artifacts became more detailed and concrete. After each iteration, the participants and I reflected on the artifacts that were created 78
as outputs of the activity. These reflections often illuminated aspects of the social system that had not been considered earlier, as the artifact became more detailed. This process was important because it made the initial implementation of the technology probe grounded in real tasks. In addition, it gave a richer awareness of the goals, concerns, and social context that were initially studied in the Phase 2 interviews. A summary of the design and assessment activities that took place before the probe use period is shown in Table 6.4. Activity
Description
1. Develop & Review Remote Communication Tasks
I developed tasks based on Phase 2 interview data. Each family reviewed tasks and made corrections, added details, and described additional tasks that were not mentioned in the interview study.
2. Create Paper-Based Mockups
Each family sketched ideas of a phone interface that fit the tasks and abilities of the young adult user. The activity used paper and colored pens. After doing the sketches, the family and I discussed the potential designs and ideas.
3. Review High-Fidelity Mockups and Flow
Each family reviewed higher fidelity mockups and the user interface flow I created. When possible, I presented multiple design alternatives to encourage criticism and discussion.
4. Evaluate Technical Implementation (Usability Test)
After the design was implemented on the mobile device, the daughter participated in a usability evaluation of the device. I ensured that the participant could complete all the target tasks with the device.
Table 6.4: Design and assessment activities before probe use period 6.6.1. User Task Development and Review This research employed task-centered design [Lewis & Rieman, 1993] to guide the UI design of the system. In task-centered design, designers create a suite of concrete, detailed descriptions of user tasks to determine the requirements of a system. I created a first version of tasks for each family based on the data collected in the Phase 2 interviews. The tasks represented descriptions and goals for current remote communication activities that the family is doing, as well as goals for future activities that the family wishes to do, but cannot do today. I then interviewed the families again and reviewed the tasks, gathering additional detail, corrections, constraints and social factors affecting the tasks. Figure 6.4 shows two tasks involving Linsey making calls. The first task takes place currently, and the second
79
task was mentioned in the interview as something that would be nice for Linsey to be able to do. The figure includes the additional feedback gathered when I reviewed the tasks with the family. From the feedback, it is clear that there are behavioral and social factors that need to be considered when designing to support Task 2: her parents were concerned that Linsey would call her swim teacher too much and at inappropriate times. Task 1: Linsey calls her parents because the bus doesn’t show up Linsey is at the East Boulder Rec Center and wants to catch the bus to get home. The bus hasn’t come and Linsey is wondering what’s going on. She calls home to tell her parents. Feedback Status: Frequent Dad: “does this all the time” Mom: “It gives her a reason to call. She likes to call, she needs to learn etiquette. It’s something we have to constantly reinforce. The phone is not a toy, it is a tool that needs to be used appropriately. Linsey will call our house 15 – 20 times in a night.”
Task 2: Linsey calls her swim teacher to say she isn’t coming to class Linsey knows it’s Monday and so she has swimming at 4pm. Her roommate usually helps her know when she needs to leave the house to catch the bus. Today she doesn’t feel well, so she wants to call to tell her teacher that she isn’t coming to class. Feedback Status: Never Right now, she’s not given the ability to call anyone except a select group of family members, because she’ll call too much Mom: “could call 15 times, not once”
Figure 6.4: Sample remote communication tasks with feedback These tasks served as living functional requirements for the technology. Although the probe was not intended to support every task in its fullest form, the tasks guided the probe design to focus on the most important, practical functionality. 6.6.2. Design Activity: Paper-Based Mockups In the first participatory design activity, families created paper-based sketches of the phone interface shown in Figure 6.5. The families were instructed to create at least one sketch for each task developed in the first exercise. Each family had three to five tasks, and created roughly the same number of paper sketches. Participants were 80
given cut-outs of a blank PDA, and “talk balloons,” so that their sketches would be guided by the general form factor of a PDA. The talk balloons were intended to encourage participants to consider the audio as well as visual interface. Participants were given colored pens and glue to assemble the sketches. Figure 6.5 demonstrates the equipment that the families used in this task.
Figure 6.5: Paper-based mockups for remote communication tasks Translating the tasks into sketches was not a completely natural activity for the participants. For the most part, the families did not reflect any further in this stage than they did on the tasks themselves. This activity required significant facilitation to stay “on track.” For example, the Roberts family quickly strayed from remote communication tasks that were developed from the interview and drew sketches of the device assisting the daughter in other tasks, such as grocery shopping, vacuuming the floor, and catching the bus. This activity facilitated a creative expression on the part of the parents about what they would like technology to do further in the future, but was not grounded in their current tasks. At the end of the activity, Linsey’s mother reflected on the sketches and the role they should play in the study. She said that these are ideas that they are working towards, but are not the primary goal of the 81
system, which is to support Linsey in calls. She reflected on the importance of simplicity and suggested that we keep the phone functionality very simple to start out with. Linsey was generally disengaged in this activity and did not participate in thinking about the functionality, although she did draw a pink phone (see Figure 6.5). This guided us to add a lot of pink in the user interface. This activity was also fairly difficult for the Smith family and required a lot of facilitation. Sally and Margaret participated in this activity because John was out of town. Sally did not create any sketches on her own, and seemed to lack confidence in drawing interface components. I did a lot of the drawing while discussing ideas and goals with Sally. Although her mother and I tried to include her, Margaret was disengaged in this activity and did not create any drawings or make any design suggestions. 6.6.3. High-Fidelity Mockups and User Interface Flow From the paper-based mockups, I created slightly higher fidelity mockups in Adobe Photoshop and added color, more realistic buttons, and other details in the UI. These mockups captured the design and functionality envisioned in the sketches, and were influenced by the existing design of AbleLink software. The mockups included an interface for initiating and receiving calls. For the Roberts family, the mockups included a Memo function that was designed to look and behave like the family’s existing memo recorder. I put these mockups in a PowerPoint presentation, and annotated them with comments on functionality and flow through the UI. Figure 6.6 is an example of one of the annotated mockups for the main screen of the system, which has a clock and buttons that enable the user to listen to a memo and make a call. In the annotations, I mapped the UI functionality with the goals of the caregivers and primary user as they were articulated in the interviews and task development activity.
82
Figure 6.6: Annotated color mockup
I reviewed these designs with the families in a face-to-face meeting. We changed and further annotated the designs together based on the discussion. The increased detail in the designs supported a deeper level of reflection by the parents and brought to the surface considerations that had not come up before. For example, the issue of privacy emerged for the Smith family at this point. The father reflected on whether the phone would have more personal information on it than Margaret’s current technology, which could compromise her privacy if the device was lost or stolen. We decided that it probably would not, but we would continue to consider this as the design evolved. For the Roberts family, the designs helped the family further reflect on how the system could mediate appropriate social behavior. They thought of audio prompts to remind Linsey to conduct phone calls in a socially appropriate way such as, “Linsey, are you sure you need to call Mom at work?”
83
6.6.4. Initial Implementation and Usability Evaluation on the PDA From the feedback on the mockups, I implemented the initial functionality on the PDA-based mobile phone. The implementation of each of the probes is described in detail in Chapter 7. For the phone contact list, I collected photographs from the families and took additional photographs. I recorded audio prompts for each of the phone contacts for an outgoing and incoming call. I introduced the implementation of the designs to the family and conducted a usability evaluation. This was the first time the families saw the phones. We conducted a separate usability session where each daughter performed the tasks that the phone supported: sending a call, receiving a call, and in the case of the Roberts family, listening to a memo. In both cases, the usability sessions were successful and the individuals were able to complete the tasks without significant problems. I attribute this success to two factors: the parents played an instrumental role in designing the UI, and the phone had very few features. After the successful usability evaluation we began the technology probe study. If we had discovered usability problems during this session, we would have performed another iteration of design updates and evaluation. 6.7. Activities during the Technology Probe Use Period During the eight-week technology probe study, the families participated in various research activities and the daughters used the technology probe in place of their previous mobile phones. The activities included a voicemail diary, design and reflection meetings, technology updates, controlled usability evaluations, and a phone call journal. I also conducted participant observation while spending time with the family on outings, and while riding the bus with the probe users. Table 6.5 summarizes the activities that took place as well as the data collected during the use period of the probe study.
84
Activity
Description & Data Collected
1. Young adult with cognitive disabilities used remote communication system for daily tasks
Participant used the probe for 8 weeks (longer if the family chose to continue using the system).
2. Daily voicemail diary
Participants left a daily message on a dedicated voicemail answering three questions: describe phone usage that day, describe any problems, and describe any new ideas. Data Collected: Qualitative data about daily usage
3. Weekly design meetings
Families participated in design activities where they brainstormed about improvements for the system interface and functionality based on their daughter’s experience using the system. Data Collected: Mockups of design updates, qualitative data about problems and ideas
3. Technology updates and usability evaluation
The system was updated with modifications based on outputs of the design meeting. Usability evaluations were conducted to ensure that the modifications did not negatively impact usability. Data Collected: Evaluation of updates
4. Phone call journal
Parents maintained a paper journal recording every call with their daughter for five to seven days, towards the end of the technology probe study. Data Collected: Qualitative description of phone calls by parents, categories for purposes of phone calls
Data Collected: Extensive usage logging through the probe
Table 6.5: Activities during the technology probe use period 6.8. Daily Use of the Probe During the technology probe use period, Linsey and Margaret carried their new phones with them and used them in a naturalistic manner in a variety of settings (such as at home, waiting at the bus stop, and riding the bus). Throughout the study, the probe logged usage extensively, recording every screen tap that registered with the software. Table 6.6 demonstrates an excerpt of the log file generated by the probe. In this sequence of log entries Linsey makes a call to her mom’s mobile phone and an hour later receives a call from her mom’s mobile phone. (A more detailed description and example of the probe log data is given in Chapter 7.)
85
Activity
Description & Data Collected
10/24/2006 10:11:02
Phone Book: Pressed picture to get to call screen. Contact: Mom Cell
10/24/2006 10:11:05
Call Screen: Pressed contact picture to send call to: Mom Cell
10/24/2006 10:12:02
Call Screen: Call terminated by contact
10/24/2006 10:12:02
Call Screen: Pressed oops/stop button to end call
10/24/2006 10:12:09
Phone Book: Back button clicked to Discovery Desktop
10/24/2006 11:21:18
Call Screen: Receiving Call from Mom Cell
10/24/2006 11:21:31
Call Screen: Pressed oops/stop button to end call
Table 6.6: Excerpt from the probe log The usage logs were an important source of data, as they captured aspects of use that were not observed through any of the other data collection methods. For example, through the logs we found that Linsey sometimes made unintentional calls by tapping the screen when the backlight was off. The log data showed that she would sometimes press a contact’s picture once to get to the call screen (which would prompt her to tap the picture again to make the call), and then leave the phone for a few hours. She would then accidentally make the call by tapping the screen to turn the backlight back on. This observation led to the addition of a “time-out” feature where the phone would automatically return to the main screen if a contact picture was pressed but the call was not completed after 30 seconds. 6.8.1. Voicemail Diary Families kept a nightly voicemail-based diary throughout the eight weeks of the study. Two benefits of voicemail diaries over a paper-based diary are that participants can record a new message simply by dialing a phone number without having to keep a physical diary with them at all times, and I can monitor the messages remotely [Palen & Salzman, 2002]. I listened to and transcribed the voicemails on a nightly basis to follow up on any reported issues and occasionally give positive reinforcement for leaving the diary messages. Small laminated cards with diary entry instructions were given to the participants to be kept by the phone or with the participant. The families were asked to call a designated voicemail box each evening and leave a message answering the following three questions: 86
•
Briefly describe any calls made today
•
Describe any problems with the phone or say “no problems”
•
Describe ideas for improvement or new features for the phone or say “no new ideas”
The diary messages were left by either a parent, companion caregiver, or the young adult with cognitive disabilities. In the Roberts family, the voicemail diary was left by Linsey’s roommate during the week, and by Linsey’s mother on the weekends or when there were problems with the phone. In the Smith family, the voicemail diary was left sometimes by Margaret’s dad and sometimes by Margaret (with some coaching by her dad). Margaret’s parents suggested that Margaret leave a voicemail diary as a way to have more responsibility during the study. The voicemail diary messages provided useful data about how the probe was used and captured descriptions of problems soon after they happened, with details that the caregivers may have otherwise forgotten. This data was discussed in weekly face-to-face design and reflection meetings. 6.8.2. Design and Reflection Meetings During the probe use period I met with each family once a week in their homes. The purpose of the design and reflection meetings was to look at the usage data of the technology probe together with the family, discuss anything that had come up in the voicemail diary, and conduct participatory design activities for new features. In the original study design, I planned to have a reflection meeting every week and a design meeting every two weeks. It turned out that it was much more effective to be less structured and more “opportunistic” about the participatory design activities and conduct meetings when there were features that needed to be redesigned or added. For the participatory design activities, I brought paper print-outs of the current UI of the probe, along with blank phone cutouts and audio call-outs and colored pens (like the materials used in the first design activity). In some cases I brought design mockups that demonstrated ideas of potential features that had been discussed. During the meeting, we would first discuss changes or new features and look at
87
supporting data. We would then create some sketches for the changes or evaluate mockup(s) that I had created. Any design changes to the device that emerged from the participatory design sessions were implemented and added to the device within one to two weeks. The short turnaround time created a practical limitation on what types of ideas could be pursued as implemented features during the study. The ideas that emerged but were out of scope were still captured. 6.8.3. Technology Updates and Controlled Usability Evaluations Immediately following each update to the system I conducted a usability session to evaluate the changes with the participants with cognitive disabilities. The usability session generally included having the user complete the tasks that were affected by the change. These sessions were controlled in the sense that the tasks were conducted “by request” during the meeting, rather than arising naturally. These sessions also provided training on newly added features. Because the technology probe changed slowly over the course of the study, these feature updates tended to be small and isolated. The usability sessions were always successful, demonstrating that the user was able to use the new functionality (although not all of the new features were incorporated by the user into regular use, as I discuss in Chapter 8). 6.8.4. Phone Call Journal The idea of the phone call journal emerged during the technology probe study with the first family. It became evident that there was a large amount of quantitative and qualitative data available describing the phone usage to support hypothesis generation about the purpose of the phone calls. However, there was no data directly about the phone calls (the voicemail diaries tended to be summative, given that they were recorded once a day). I observed in the log data that there were patterns in the phone calls at particular times of day. I wanted to get a better idea of the purpose of those calls and how they related to the families’ activities.
88
The phone call journal activity was focused on understanding the nature and purpose of the calls made each day. It took place six weeks into the technology probe study after a significant amount of data was available and calling patterns had developed. This allowed me to explore if the purposes of the users’ calls had qualitatively changed since the beginning of the study. The activity had three components: an initial meeting with the caregivers, a weeklong journal study, and a follow-up meeting. In the initial meeting, the caregivers and I collaboratively created categories for the purpose of their daughter’s phone calls, based on recent concrete examples. During the journal activity, each caregiver kept a small paper-based journal where they logged and categorized as many incoming and outgoing calls with the daughter as they could, based on the categories we developed in the initial meeting. In the follow-up meeting we discussed the journals and modified the categories based on the data. This activity is similar to the Experimental Sampling Method (ESM) as used in [Palen & Salzman, 2002], but the journal entries were driven by the occurrence of specific events (phone calls) rather than random time samples. An entry page for the phone call journal for the Roberts Family is shown in Figure 6.7.
Figure 6.7: Phone call journal entry of a call Linsey made to her mother 6.9. Activities after the Technology Probe Use Period At the end of the technology probe use period, I conducted a brief closing interview to discuss any remaining usage issues and discuss how communication
89
tasks changed over the course of the usage period. During the probe study I developed a close relationship with each of the families and we have remained in contact socially. At the end of the probe study, the families were given the option to continue using the new remote communication system or return to the previous mobile phone. If they chose to continue using the new system, they would receive the software and hardware for free and the phone would be transferred over to the families’ personal service plan. AbleLink Technologies agreed to provide technical support to the families after the end of the study. Both families chose to keep the new phone and are still using it currently (which makes nine months of usage for the Roberts family, and five months of usage for the Smith family). 6.10. Summary In this chapter I provide a detailed description of the probe study activities and artifacts created. The goal of the probe study was to create an environment to support design-in-use, in which families can participate as co-designers and requirements emerge through use. During the study, I triangulated quantitative and qualitative data to gain multiple perspectives on how the participants used the probe. At the end of the study, the probe had evolved with each family, although many of the features were the same between the two families’ systems. In Chapter 7, I describe the technical implementation of the system. In Chapters 8 and 9, I discuss how this study informed an understanding of a socio-technical environment composed of caregivers, care recipient, and technology. In Chapter 10 I discuss the unique affordances of an evolving technology probe, and suggest that this method can be effective in understanding the requirements for a meta-design environment.
90
Knowing is not enough; we must apply! – Goethe
7
Implementation, Evolution, and Evaluation of the Probe 7.1. Introduction In this chapter I describe the implementation, evolution, and evaluation findings of the two technology probes. The platform was a PDA-based mobile phone and the software was built on applications designed and implemented by AbleLink Technologies [AbleLink Technologies, 2007]. First, I describe AbleLink’s three mobile applications that were used in this study: Pocket Ace, Discovery Desktop, and Schedule Assistant. I modified the UI and added functionality to the AbleLink software so that it would meet the goals of the technology probe as it evolved through the experiences of the participants through the course of the study. I describe the addition of usage logging that was added to support the probe logging requirements. I then describe the modifications that were made in the software for each family, in chronological order as they were implemented through the course of the probe study. I conclude by summarizing the usability findings during the probe study, presented as design implications for PDA-based assistive technology for users with cognitive disabilities. 91
7.2. Mobile Technology Platform The technology probe was implemented on a Verizon XV6700 (also known as UT-Starcom PPC-6700 and HTC Apache), and is shown in Figure 7.1. The hardware was chosen based on a number of criteria. First, the XV6700 has a good form factor: it is fairly small and light but thick and easy to hold on to and has a relatively large and high resolution color touchscreen (1 ¾ inches wide x 2 ¼ inches tall, 240 x 320 pixels). As I did not intend to use any of the hardware buttons or the slide-out QUERTY keyboard, they needed to be unobtrusive and easy to hide.
Figure 7.1: Verizon XV6700 phone A second reason for choosing this phone was that it was offered by Verizon. Both of the families who participated in the probe study already had family plans with Verizon. I did not want this to be a barrier for adoption of the phone at the end of the study, knowing that I could not expect the family to change cellular service providers after the study. Finally, this device was chosen because it used the Windows Mobile operating system on which all of AbleLink’s software ran. I knew the phone would need a protective case as it would be used in many environments. The case needed to provide access to the touchscreen, hide the buttons and slide-out keyboard, and be easy to hold on to. This unexpectedly turned out to be a challenging set of requirements, as most cases for this model exposed the buttons 92
and keyboard. My first attempt was an adapted iPod neoprene case. This was used for the first 3 weeks with the Roberts family. The main problem with this case was that it had two Velcro straps on top that Linsey would play with and when opened they exposed the power button and storage card. This led to Linsey frequently turning off the phone or popping out the storage card by accident. The second case was a rubberized “flexiskin” that worked much better: it was designed to permanently stay on the phone In addition, it covered the power button, storage card, and other hardware buttons. As a bonus, this case came in multiple colors so the participants could choose their preferred color (Linsey chose pink and Margaret chose blue). This type of case was not made for the XV6700 model so I had to buy a different model and modify it by cutting holes in the right places for the microphone and antenna. Figure 7.2 shows the two cases. A great deal of time was spent in the first three weeks of the Roberts’ probe study trying to find ways to prevent Linsey from unintentionally turning off the device or popping out the storage card – both platformlevel usability issues that were not identified in the controlled usability evaluations we conducted during the family meetings.
93
Figure 7.2: Two cases for the phone: (A) a neoprene case designed for an iPod, and (B) a rubberized "flexiskin" case designed for a different phone model 7.3. AbleLink Software The remote communication system is based on software from AbleLink Technologies. The collaboration agreement gave me full access to AbleLink’s software source code and my own branch in their source code repository, so I was able to modify and augment the code freely. The source code is a combination of Visual Basic .NET and C#, using the .NET Compact Framework running on the Windows Mobile 5 operating system. AbleLink has a number of applications written for this environment; the three applications used in this study were a mobile phone application called Pocket Ace, an application launcher called Discovery Desktop, and a scheduling/alarm application called Schedule Assistant. 7.3.1. Theme Framework AbleLink has a theme and configuration framework that supports some customizations out of the box. Each application has an initialization file that specifies 94
the theme and configuration options for the given deployment. The theme framework supports changing image and sound files used in the interface, and the configuration options control some application functionality. I was able to utilize this configuration and theme framework for most of my customizations for the probe participants. When I added and modified features, I extended the framework as much as possible so the features could become configuration options. This largely enabled me to switch between participant configurations without having to recompile code, which was very useful during testing and debugging, and for code maintenance in the future. When I implemented the Memo Player, I created a configuration file that used the same format as the other theme files (as detailed in Figure 7.3).
Figure 7.3: Configuration file for customizing the Memo Player application Through this configuration file one can dynamically change the background color and other look-and-feel aspects of the user interface, the maximum audio length, and the behavior of navigation buttons (“discoveryDesktopMode”). The configuration parameters that were expected to change often (the selected audio and selected image) were configurable through a caregiver’s interface on the PDA. 7.3.2. AbleLink Applications Used in the Study I will describe the existing functionality of the AbleLink software that was used in the technology probe study. Since AbleLink is continually creating new versions of
95
their software, this description is accurate for the versions of their software available in the summer of 2006. All of AbleLink’s mobile applications are based on a touchscreen and do not use any hardware buttons on the devices. The software disables the hardware buttons, so pressing them by accident has no effect. Limiting input to a touchscreen has a few benefits: first, hardware devices change over time and are different across cellular phone providers. Hardware buttons across devices have different functions and locations, so relying on these buttons would potentially require users to re-learn the application when they upgrade their hardware. For example, my participants use the Verizon Wireless network, so the hardware I used (the VX6700) was different than AbleLink’s test device (the i-mate JAM) which works on other cellular networks. Because both devices run on the Windows Mobile operating system, we were able to transfer the software between the devices without making any modifications. Another benefit of touchscreen-based buttons is that they can be customized with respect to size, color, and response behavior (such as timing out to avoid being clicked twice accidentally). AbleLink’s mobile applications all have a “button talk” feature that enables each button to announce its functionality when pressed, and execute that functionality when pressed a second time (e.g., a button talk message for a button in the phone contact list might be “To call Melissa, press the picture”). 7.3.2.1. Pocket ACE Pocket ACE is AbleLink’s mobile phone application that runs on the Windows Mobile platform. Pocket ACE is shown as (A) in Figure 7.4. Pocket ACE has the following features: •
Picture-based calling, using a contact list that displays four contacts per page
•
Custom audio recording for each person in the call list, both for outgoing and incoming calls (e.g., “To call Melissa, press the picture,” and “Melissa is calling! Press the picture to answer the call”)
•
Recognition of incoming calls from someone in the contact list by displaying their picture as well as playing the custom audio recording 96
•
Optional touchscreen number pad to dial any phone number
•
An administration interface for caregivers to add and remove contacts, associate images, record audio messages for each contact, and set other configuration options
Figure 7.4: AbleLink's software used in the probe study: (A) Pocket Ace, (B) Discovery Desktop, (C) Schedule Assistant 7.3.2.2. Discovery Desktop Discovery Desktop is AbleLink’s version of Windows Desktop: it displays shortcuts to launch other applications and is the “main screen” that the user sees when starting up the device. Discovery Desktop is shown as (B) in Figure 7.4. Discovery Desktop has the following functionality: •
Icon-based buttons to launch applications
•
A custom audio recording for each button (e.g., “To make a call, press the picture”)
•
Date and time display
•
Notification of upcoming events, if Schedule Assistant is used (described below)
97
•
An administration interface for caregivers to set configuration options such as adding, reordering, and removing application shortcuts, recording audio messages, and associating images with applications
7.3.2.3. Schedule Assistant Schedule Assistant is an application that provides time-based reminders through images and audio messages. Schedule Assistant is shown as (C) in Figure 7.4. Schedule Assistant has the following features: •
A list display of upcoming events (per day) with icons and an “hourglass” display to represent proximity of each event to the current time
•
An audio and visual alarm that announces an event
•
A custom audio message that can be played when the alarm is acknowledged (e.g., “It is time to go to work. Don’t forget to bring your keys and bus pass”)
•
An administration interface for caregivers to add and update events, associate images and record audio messages for each shortcut, and set other configuration options
These three applications formed the foundation of the technology probe implementation. These applications were designed to support custom pictures and audio messages, which made it very conducive to do these types of customizations for each family. I will now describe the modifications that I made to the AbleLink software. 7.4. Usage Logging I created a framework for usage logging for the technology probe that allowed me to collect very detailed usage data. The logging functionality was written in C# in a shared library that was available to all applications. The generated logs were CSV files, a standard log file format preferable to XML because it does not rely on opening and closing tags, is more concise, and is less likely to become corrupted. (A major problem with XML-based log files is that if the application creates malformed XML,
98
e.g., fails to write out a closing tag due to a power failure, the parser can no longer read or update the file and it must be discarded or manually repaired by a developer.) The logging framework has four logging categories, and a developer can choose which categories to include in a given deployment. The logging categories are shown in Table 7.1. For the technology probe study, we used all four logging categories. Logging Category
Description
Trace
Every time a UI control is activated (e.g., every button press and screen tap on a form element) The initiation of some application function (e.g., making or accepting a call, or launching Pocket Ace from Discovery Desktop) (The difference between Trace and Activity log entries are that Trace entries do not initiate an application function.) Any activity in the administration interface System-level operations (e.g., the starting and stopping of applications)
Activity
Admin Activity System
Table 7.1: Four Categories of Usage Logging Each of the categories had two levels of log messages: Info and Error. As the label suggests, the Info messages were logging normal activity and Error log messages reported error states. Each log entry has five elements: a timestamp, log category, log level, reporting application, and message. I updated each application to create appropriate log entries during execution. Table 7.2 is a sample of log entries created by Pocket Ace. In the following log entries, Linsey makes a call to her mom’s mobile phone, receives a call from her mom’s mobile phone an hour later, calls her house, and then navigates through the phone book. Linsey’s probe uses “button talk,” so the Trace entries log the first time a caller button is pressed and announces its function, (e.g., “To call mom’s cell, press the picture”), and the Activity entries log when the caller button is pressed a second time to make the call.
99
Timestamp
Category
Level
Application
Message
10/24/2006 10:11:02
Trace
Info
PocketAce
Phone Book: Pressed picture to get to call screen. Contact: Mom Cell
10/24/2006 10:11:05
Activity
Info
PocketAce
Call Screen: Pressed contact picture to send call to: Mom Cell
10/24/2006 10:12:02
Activity
Info
PocketAce
Call Screen: Call terminated by contact
10/24/2006 10:12:02
Activity
Info
PocketAce
Call Screen: Pressed oops/stop button to end call
10/24/2006 10:12:09
Trace
Info
PocketAce
Phone Book: Back button clicked to Discovery Desktop
10/24/2006 11:21:18
Activity
Info
PocketAce
Call Screen: Receiving Call from Mom Cell
10/24/2006 11:21:31
Activity
Info
PocketAce
Call Screen: Pressed oops/stop button to end call
10/24/2006 11:21:33
Trace
Info
PocketAce
Phone Book: Pressed picture to call contact: My House
10/24/2006 11:21:34
Activity
Info
PocketAce
Call Screen: Pressed contact picture to send call to: My House
10/24/2006 11:21:57
Activity
Info
PocketAce
Call Screen: Pressed oops/stop button to end call
10/24/2006 11:22:01
Trace
Info
PocketAce
Phone Book: Next button clicked
Table 7.2: Log file entries showing recipient and duration of phone calls Through this extensive usage logging I was able to reconstruct usage of each application and understand how users arrived in error states that were never reported via other data; in general this data served as a very important complement to the other more qualitative data collected through the study. 7.5. Data Aggregation and Visualization Tool During the probe study I quickly became overwhelmed with data. I needed a way to aggregate and visualize the various qualitative and quantitative data sources, to be able to generate and evaluate hypotheses. I had thousands of entries from the probe log, transcripts from daily voicemail diaries and memo recordings, notes from meetings, and information about the family’s schedule. I imported the data into an MS Access Database, created queries that processed the log data into meaningful units (e.g., completed phone calls), and then created a form to view the data a day’s worth at a time. I used this view to develop a high-level understanding of the probe usage. The caregivers also viewed the form during design meetings and found it
100
useful to gain insights into how their daughters used the phone. The form created for Linsey’s probe is shown in Figure 7.5. The form for Margaret’s probe was slightly different and displayed data about the features on her probe.
Figure 7.5: Researcher tool developed to view and analyze quantitative and qualitative data collected during the probe study 7.6. Linsey’s Probe I now describe Linsey’s probe as it was initially implemented and as it evolved during the probe study. 7.6.1. Initial Functionality Linsey’s probe had two applications: the mobile phone application (a modified version of Pocket Ace), and the Memo Player. The probe used Discovery Desktop as the main screen for Linsey to launch these two applications. Linsey’s version of Pocket Ace had a blue background and pink buttons (chosen by the family). It did not include a number pad dialer, as Linsey was not able to recognize numbers consistently. The phone started out with eight picture-based call contacts: 1) her parents’ home phone, 2) her mom’s mobile phone, 3) 911, 4) her apartment, 5) her mom’s work phone, 6) her dad’s mobile phone, and 7) her roommate Abby’s mobile
101
phone. Linsey’s version of Discovery Desktop and Pocket Ace are shown in Figure 7.6. Discovery Desktop had two icons on it: one to launch the Memo Player, and one to launch Pocket Ace.
Figure 7.6: Linsey's version of Discovery Desktop (left) and Pocket Ace (right) 7.6.1.1. Memo Player Application The Memo Player was a custom-written application for Linsey’s probe that played and recorded a single audio message. It was very similar in functionality to an inexpensive stand-alone memo recorder device that one might find at Radio Shack. The Memo Player was intentionally modeled after a memo recorder that Linsey’s mother used to record a daily message for Linsey. The software version was designed to match the stand-alone device’s simplicity and ease-of-use. The UI had three buttons: a large play button which had an image on it, a small record button, and a navigation button to return back to Discovery Desktop. The Memo Player had a caregiver interface to record a new memo or choose from previously recorded memos (marked with a time stamp), and update the picture on the Play button. Figure 7.7 shows the play and record interfaces of the Memo Player.
102
Figure 7.7: Linsey’s Memo Player showing the Play screen (A) and the caregiver’s Update Memo screen (B) The goal of the Memo Player was to give Linsey a daily message describing her day. The usage of the memo is described in more detail in Chapter 8. 7.6.1.2. Navigation between Applications AbleLink had not designed a way to navigate between applications, such as navigating back to Discovery Desktop from Pocket Ace or the Memo Player. I modified Pocket Ace so that when the user was on the first page of the contacts list and pressed the left “previous page” button, it returned to Discovery Desktop. I used the same “previous page” button in the Memo Player application and Margaret’s Bus Arrival application (described below) to return to Discovery Desktop. Although this meant that the previous page button in Pocket Ace could result in two possible actions (either navigating to the previous page of contacts, or returning to Discovery Desktop), it provided a consistent behavior that the “previous page” button would always get the user back to her home starting point, Discovery Desktop. The two users did not have trouble understanding and using the previous page buttons to move between applications. 7.6.2. Evolutionary Changes Linsey’s probe evolved on different levels, from the accessories (such as the case and battery), the operating system level, the software functionality level, and finally at the UI level. Linsey’s probe evolved due to breakdowns as well as new needs and 103
ideas. As the Roberts family was the first to participate in the probe trial, they encountered the majority of the stability-related issues. In addition to changes that could be considered customizations, Linsey’s probe also evolved due to bugs found in the software and unforeseen problems with the platform. Table 7.3 summarizes the evolutionary changes of Linsey’s probe during the usage period. Week
Evolutionary Changes
Week 1
• Fixed the case (with tape and a rubber sticker) so that it would be harder to pop out the storage card and press the power button
• Changed the call contact picture for Linsey’s parents from a picture of them to a picture of the house (so Linsey could distinguish between their home and cell numbers)
Week 2
• Added accessible launcher to Windows Mobile screen so Linsey could restart the phone
• Carin found a better rubber sticker to cover storage card Week 3
• Added a new call contact (a friend of Linsey’s) • Changed the main picture of the memo recorder to look more like a calendar
Week 4
• Updated incoming call message so it repeats over and over • Changed Memo Player so pressing memo button multiple times does not cause it to get "backed up" (by changing from synchronous to asynchronous)
• Trained Carin on how to do a “soft reset” on the phone when memory is low Week 5
• Added 6 new call contacts (extended family in New York) • Added ability to “disable” a contact temporarily, and disabled Mom’s work number. • Removed volume control at the top of the screen so it could not get hit accidentally • Trained Carin on how to add contacts
Week 6
• Changed phone case to flexi-skin • Added button talk reset, so phone dialer would return to phone book after time-out
Week 7
• Added screen lock • Added clock talk to Discovery Desktop so it would tell Linsey the date and time
Week 8
• Replaced the battery with an extended battery to improve phone life
Week 9
• Carin added 3 new contacts by herself • Created a troubleshooting guide for caregivers to help Linsey recover from errors
period
with the phone
Week 10
• Updated Memo Player so it only kept the last 10 recorded memos (since Carin and Abby decided they did not need to use older ones)
Table 7.3: Evolution of Linsey's phone Most of the breakdowns occurred at the platform level. The first critical issue was that Linsey was frequently pressing the power button on the phone or popping out the storage card by accident. Carin and I searched together for a better case for the probe and tried different methods of preventing Linsey from accessing these areas. This was
104
a familiar problem for Carin, as she was used to making simple modifications to Linsey’s technology – using inexpensive household items – to make the technology more accessible and robust. For example, Carin put a dot of nail polish on one side of the phone’s USB charger input, and a matching dot on the side of the phone, so that Linsey would turn the input the right way when plugging it in. It was very beneficial to collaborate with Carin and leverage her expertise when addressing these problems. Ultimately, the flexiskin case in combination with a rubber sticker that Carin attached over the storage card fixed the problem. Another common category of problems were between-task breakdowns, which would have been very hard to find in a controlled task-based usability evaluation. These included needing to re-start the software after the battery ran out, accidentally executing a function when the backlight was dark, and accidentally tapping the touchscreen while carrying the device in a purse or bag. A significant part of the changes on Linsey’s probe were software modifications to compensate for platformlevel breakdowns and to handle between-task breakdowns. Three examples are the accessible launcher, the button-talk reset feature, and the screen lock. 7.6.2.1. Accessible Launcher from Windows Mobile Today Screen By design, users of AbleLink mobile software are never supposed to see the underlying Windows Mobile operating system screens, such as the main “Today” screen. However, due to platform-level breakdowns, this actually happened fairly frequently. Although there was a function to re-launch the AbleLink software when the system rebooted or recovered from a dead battery, the Windows Mobile operating system rarely did this correctly. When the user reached the Today screen she became stuck, because the only way to re-launch the AbleLink applications was to navigate through a Windows Start Menu and find the application listed there. This was too complex for our users to do, so they were unable to use the phone until a caregiver or researcher re-launched the applications for them. To compensate for the platform and operating system level problems, I added a large heart-shaped button in the middle of the Today screen and hid most of the other controls displayed on this screen. This enabled users to re-launch the AbleLink
105
software themselves. The modified Today screen is shown in Figure 7.8. The probe users easily learned the effect of pushing this button and were able to recover from many more breakdowns without relying on a caregiver. The button had a “heart” icon to follow the metaphor that evolved in the Roberts family of the probe as Linsey’s friend – it had even been nicknamed “Buddy.” When we added the heart button, we explained to Linsey that she pressed Buddy’s heart to bring him back. We began Margaret’s probe study with the accessible launcher.
Figure 7.8: Modified "Today" screen with accessible launcher (heart button) 7.6.2.2. Button Talk Reset As described earlier, the buttons used a mode that AbleLink calls “button talk,” which means that a single press announces the function of the button, and a second press activates the function. In the original software, there was no time component between the first and second button presses (in other words, the second press would activate the function whether 5 seconds or 5 hours had gone by). During Linsey’s probe study, log data showed that Linsey would occasionally only press a contact name once, and then leave the phone alone for a long period of time. The screen backlight goes off after a few minutes, so on a few of these occasions Linsey would send a call by pressing the screen once when the backlight was off (which could not have been intentional). This was found in the logs when Linsey unintentionally called 106
911 in this way. To address this, we changed the button talk functionality to reset after 30 seconds if the second button press did not take place. 7.6.2.3. Screen Lock The purpose of the screen lock was to prevent unintended taps on the touchscreen while the phone was in a purse or backpack. The screen lock hides all the buttons on the screen of Discovery Desktop, except one small button to unlock the screen. The screen lock emerged through a request by Linsey to be able to “turn Buddy off” when she carried it in her purse, because she was accidentally tapping the screen. The screen lock interface was co-designed with Linsey’s family. We stayed with the Buddy metaphor and used a heart for the lock/unlock button to be consistent with the heart used in the accessible launcher described above. Figure 7.9 shows the screen lock button on Discovery Desktop, and the screen when it is locked. We taught Linsey that pressing the lock button was “putting Buddy to bed,” and pressing the heart brought Buddy back (just like with the launcher). The screen automatically unlocks when an incoming call is received, or a scheduled alarm goes off. Linsey did not have trouble learning this feature and used it judiciously. We began Margaret’s probe study with the screen lock functionality. She also did not have a problem learning to use it.
Figure 7.9: Screen lock feature showing (A) the unlocked state, (B) Linsey’s version and (C) Margaret's version of the screen lock
107
Not all of the changes to Linsey’s probe resulted from complete breakdowns (thankfully, as this is a stressful way for both researcher and participants to design!). Through experience in usage during the probe study, the family and I co-designed new functionality and made improvements to existing functionality. These included adding an audio-based clock to Discovery Desktop, disabling phone contacts, and improving the incoming call message. 7.6.2.4. Added “Clock Talk” to Discovery Desktop AbleLink had implemented a configurable feature to their Schedule Assistant application that changed the time and calendar display into buttons that would announce the time and date in an audio message. This feature was not available in Discovery Desktop, where the time was just a visual display and not a button. In the course of the probe study, we found that Linsey – who is unable to read clocks and does not understand time or numbers very well – frequently makes phone calls to ask what time it is. To explore how technology might assist Linsey in understanding time, we added the “clock talk” feature to Discovery Desktop to see if the ability to hear the time would be useful and would reduce the number of times she made calls. The results are described in detail Chapters 8 and 9, but in summary Linsey very rarely used clock talk compared to how she used the other features of the probe. Adding clock talk did not reduce the number of times she called her caregivers. 7.6.2.5. Ability to Disable Contacts in Pocket Ace The need to disable phone contacts in Pocket Ace emerged from two different conditions. First, the Roberts family went on vacation with extended family during the probe study. While on vacation, Linsey’s mother added additional extended family as contacts on Linsey’s phone. She wanted to remove the phone contacts that would not be used while on vacation (such as the mother’s work number). Since this was only a temporary removal, it was more natural to hide or disable these contacts rather than remove them. The second motivation of this requirement was more social in nature: since Linsey had the tendency to make too many calls, Linsey’s mom
108
wanted to be able to hide and show different contacts on her phone at different times, in a sense to give Linsey’s contacts “a break” from Linsey’s calls. 7.6.2.6. Improved Announcements for Incoming Calls In the original AbleLink implementation of Pocket Ace, an incoming call was announced by playing the incoming audio message a single time, e.g., “Melissa is calling, press the picture to answer the call.” Linsey’s parents, confirmed by the probe logging data, said that Linsey very rarely answered incoming calls, and the family and I guessed that it may be because she does not hear this announcement. To address this, we added a distinct alert sound before the audio announcement, and we looped the alert and message over and over until the call was answered or went to voicemail. After this change, Linsey answered incoming calls more often. We began Margaret’s probe study with this functionality. 7.6.2.7. Additions and Changes to Phone Contacts Over the course of the probe study, Linsey’s mother incrementally added more contacts to her phone. At the beginning of the probe study she had seven contacts, and by the end of the study she had 16. She called 12 of her contacts on a regular basis at least twice a week. Most of her contacts were extended family members, as well as two friends who were close to her age. As mentioned above, Carin would periodically disable contacts in Linsey’s phone to give them a “break” from calls and to make sure Linsey did not go over her monthly minute allowance (once she transitioned to the family service plan). 7.6.2.8. Training and Documentation Another evolutionary change through the probe study was incremental training for the caregiver on how to update the call contacts and how to recover from problems. By the end of the probe study, Carin was comfortable adding new phone contacts to Linsey’s phone. She and I also worked together to develop a short troubleshooting guide that described how to diagnose and recover from common errors on the phone (such as the power button getting accidentally tapped or a dead battery). This guide
109
was Carin’s idea to enable Linsey’s other caregivers (such as her manager at work) to help Linsey recover from problems on the phone. 7.7. Margaret’s Probe Margaret’s phone started out with slightly different initial functionality, and like Linsey’s phone it evolved through use. The probe study with the Smith family was conducted after the Roberts family probe study ended, so Margaret’s initial probe included bug fixes and basic improvements (such as the flexiskin case, accessible launcher, screen lock, and improved incoming call announcements) that were identified during the first probe study. In this sense, when Margaret started out she was already using a “second generation” version of the probe. This likely contributed to the fact that Margaret had fewer breakdowns at the beginning of the probe study. 7.7.1. Initial Functionality Margaret’s probe was initially designed to have a phone and a morning alarm. Similar to Linsey’s probe, Margaret’s phone had picture-based call contacts. Since Margaret was able to recognize numbers and learn phone numbers, the phone included a number pad screen, shown in Figure 7.10, (A), where Margaret could call by dialing. In fact her parents wanted Margaret to continue dialing the important numbers she had memorized (their mobile phones and home line), so they asked that they not be displayed as picture contacts. Margaret’s phone began with four phone contacts: her sister, brother, CallnRide (bus driver), and 911.
Figure 7.10: The number pad dialer (A) and morning alarm screen (B) on Margaret’s probe 110
The morning alarm, shown in Figure 7.10, (B), played one of Margaret’s favorite Patsy Cline songs. The alarm sounded three mornings a week when Margaret had activities and included custom audio messages describing the activity for that day (the usage of the alarms is described in more detail in Chapter 8). 7.7.2. Evolutionary Changes Margaret’s probe evolved along various dimensions during the probe usage period. New features were co-designed by the family and me, including a bus arrival information application and a missed call screen. We added additional call contacts to the phone and changed some of the pictures. For example, Margaret’s mom designed a new 911 picture that had the numbers “911,” which she thought would work better for Margaret. During the study we identified the need for additional peripherals (namely phone chargers) to support the usage patterns as Margaret moved between her parents’ home and her apartment. The evolutionary changes of Margaret’s probe are summarized in Table 7.4. Week
Evolutionary Changes
Week 1
• Added two additional call contacts (Donna and Andrea)
Week 2
• Added the bus arrival information application • Added the voicemail diary study number as a call contact
Week 3
• Fixed Accessible Launcher to also launch Schedule Assistant automatically (rather than having Margaret do it manually)
Week 4
• Added five new call contacts (extended family)
Week 5
• Added the missed call screen
Week 6
• Changed the picture of the 911 call contact
Week 7
• Gave two additional phone chargers to the family: one for upstairs in the family home, and one for Margaret’s apartment
Week 8
• Added two new alarms to the phone for Margaret’s favorite television shows
Table 7.4: Evolution of Margaret's probe The two most significant changes to the probe were the bus arrival information application and the missed call screen. These two features are described next. 7.7.2.1. Bus Arrival Information Application Margaret’s parents repeatedly expressed that an area of remote communication support that Margaret really needed was information about when the bus will arrive. 111
Margaret’s father explained that every time the bus was late, he or Sally got a call from Margaret. I created a Bus Arrival Information application to explore how providing Margaret with bus arrival information directly on her probe may change the communication patterns between Margaret and her parents. The application was a single screen that had an embedded web browser in it without any typical browser controls. The web browser connected to the internet and displayed real-time GPSbased bus arrival information for Margaret’s bus. This information was retrieved from a commercial website designed for hand-held devices [NextBus, 2007]. The screen had three buttons on it: a ‘Work’ button, a ‘Home’ button, and a navigation button to return back to Discovery Desktop. The Work and Home buttons updated the screen to display the current bus arrival information for the bus stop where Margaret waited for the bus on her way to work, or on her way home. Figure 7.11 shows an example screen of the Bus Arrival Information application. I did not modify the presentation of information from nextbus.com. It communicated arrival time in minutes (e.g., “Next Vehicle in: 5 minutes, 26 minutes, 40 minutes.”) This method of communicating time seemed to be a reasonable match for Margaret’s abilities given that she is able to read, and her mother often told Margaret to wait five more minutes if the bus had not arrived yet. The arrival information was specific to a single bus line called the “Dash,” which was identified visually through the application icon (the icon matched a larger image that was on the bus).
Figure 7.11: Margaret's bus arrival information screen launcher (A) and application screen (B)
112
7.7.2.2. Missed Calls Alert The missed call alert screen notified Margaret of a missed call (shown in Figure 7.12). This was added by request of Margaret’s parents, mainly because this feature was on her old mobile phone and they did not want her to lose any functionality when moving to the new phone. John originally designed a more complex interface for the missed call screen that displayed a list of all of Margaret’s recently missed calls, and differentiated calls that she had returned or not. He suggested that we incorporate Margaret’s missed alarms and missed calls into a single list. I encouraged the parents to begin with a simple design so that we could understand how Margaret used it first. We came up with the design shown in Figure 7.12. It is worth noting that once we added this feature, the parents did not ask again for a more complex version (the simpler version seemed to satisfy the requirement in practice).
Figure 7.12: Missed call alert screen that was added to Margaret's probe The missed call alert screen displays the picture and phone number of the missed caller. If the caller is not in the call contacts list, the screen displays the phone number with an “unknown caller” image. Pressing “Send” will initiate a call to the person shown in the picture and pressing “Done” will return the user to the phone book. 7.7.2.3. Additions and Changes to Phone Contacts Over the course of the probe study, we incrementally added more contacts to Margaret’s phone as we did with Linsey’s phone. At the beginning of the probe study she had four contacts and by the end of the study she had 12. Similar to Linsey, most
113
of her contacts were extended family members. Margaret explicitly requested to have cousins and uncles added to her phone, although she rarely or never called them (this is discussed further in Chapter 8). 7.8. Non-Functional Changes To support the technology probes and to improve performance, I made a number of changes to the code that did not affect user functionality. I added support for themes to Discovery Desktop and support for multiple phone books (contact lists) to Pocket Ace to make them more configurable. I fixed various bugs in functionality and improved performance by reducing the memory snapshot taken when the native mobile phone application was disabled by the Pocket Ace software. I reduced the quality of the audio recordings of the sound files created in the applications. The sound files were originally recorded at 16 bits, and 44 kHz. During the probe study we found that if the battery was low when the recordings were made, the recordings would have prominent “skips” or flaws in them. The Roberts family and I evaluated different audio qualities to find a balance between sound quality and performance and found that recordings at 8 bits and 22 kHz provided the best results. 7.9. Challenges of a Mobile Platform Implementing and supporting a heavily-used mobile application for eight weeks gave me a sense of the challenges of working on a mobile platform. Although the XV6700 device used in this study was cutting-edge technology and received top reviews, and Windows Mobile 5 is a popular mobile operating system, the platform was far less stable than a desktop environment. To illustrate, I will briefly summarize the hardware and operating system issues I dealt with during the probe usage study: •
Memory leak writing to log files. The operating system had a memory leak when writing to files that were stored on the storage card. This caused the applications to run out of memory and crash after approximately 24 hours. The work-around was to store the log files in memory directly.
•
Low battery caused applications to crash. When the battery was running low, the operating system would intentionally kill all the applications that
114
were running in the background. Unfortunately, this meant that the phone application would sometimes be killed and the user could not send or receive calls until recharging the device. •
Short battery life and battery depletion. The battery on the PDA-based phone did not last as long as a typical mobile phone. The battery life also deteriorated over time, so that after eight weeks the battery did not last 12 hours. To work around this, I added extended batteries to each of the probes that lasted at least 18 hours.
•
Tapping the Power button caused applications to crash. Although the Power button was only supposed to cause running applications to “hibernate,” the applications never successfully returned from hibernation and froze (and eventually crashed) when the power was turned back on.
•
Occasional volume loss. After the device had been running for a few days, the volume would often decrease or go away completely. The work-around was to perform a soft reset on the device.
•
Occasional random switch into “flight mode.” After resetting the device, sometimes it would start in “flight mode” meaning that the phone was disabled completely. The work-around for this was to reset the phone again or to manually take the phone out of flight mode.
•
Occasional loss of backlight. Sometimes the backlight would not come on when the screen was tapped. This rarely happened, and the only reliable workaround was to reset the device.
The number of problems I encountered at the platform level indicates that an important focus of AT design must be selecting and evaluating the platform. In this design environment, simplicity and robustness are paramount, and a mobile device that is “flaky” or complex to maintain will very likely be abandoned. This study has shown that it is crucial to fully understand each of the issues of a given mobile device before deploying it, so that software, training, and other strategies can compensate for the hardware or operating system shortcomings.
115
7.10. Design Implications from Usability Findings In this section I will briefly summarize the UI design implications that were found through the study, both during the controlled usability evaluations and the real-world use. These design findings are particularly relevant for mobile technology. Usability Finding
Description and Details
Challenges with touchscreen-based buttons
Touchscreen buttons lack tactile definition, which is a detriment to usability. Users did not always feel confident that a button was pressed (because of the lack of tactile response) and so pressed buttons multiple times or pressed very hard. Two design recommendations for this are 1) to play a clear, consistent sound when a button is pressed; and 2) to add a short delay after the button is pressed so it cannot be pressed multiple times in succession. Both of these should be customizable for each user.
Importance of a screen lock
When using a touchscreen on a portable device, it is very important to prevent unintentional taps on the screen while the device is being carried around. Adding a screen lock that used an imagebased button to activate and de-activate was effective for both participants in the probe study. Another approach would be a wellfitting case that includes a cover that can be easily opened and closed to protect the screen.
Simple, consistent navigation
Both users were able to navigate between multiple applications through consistently-placed “next” and “previous” buttons. Both users learned that pressing the “previous” button would always eventually return to the home screen.
Platform-level approach to usability
This chapter describes how the most challenging usability problems occurred at the hardware and platform levels, rather than in the software design. It is crucial to include the entire platform in usability evaluations.
Humanizing the technology
Among the probe users – especially Linsey – it was effective to describe the technology in terms of human behavior. For example, the screen lock was represented as putting the phone to sleep and waking it up. It seemed like the users had a natural tendency to relate to the device as a living entity.
Value of familiar pictures and custom audio recordings
The pictures and custom audio recordings were favorite features of the probe for both Linsey and Margaret. As described in Chapter 5, recognizing incoming callers with a “custom voice ring tone” may be a very useful feature for some users with disabilities to enable them to recognize that the caller is a familiar person.
Figure 7.13: Summary of usability findings through probe study
In addition to the features that were explored using the probe, two additional features emerged as having high potential value. The first was accessible voicemail. Both families asked for this, but it was too technically difficult to implement in the 116
short period of the probe study. The families wanted a simpler way for their daughters to listen to voicemail messages, without having to dial the voicemail number, enter the password, and navigate the voicemail menu. Linsey’s family in particular wanted the ability to leave memos for Linsey remotely, so that she could simply press a button on her phone and listen to the most recently recorded message. The second feature that was not directly explored was the ability for caregiver and care recipient to locate one another (e.g., through GPS data on their phones). While the idea of a caregiver locating a care recipient is fairly well-known, this study revealed that the location is important in both directions – it was important for care recipients to know where their parents were as well. 7.11. Summary In this chapter I provided a detailed description of the evolving functionality of the two technology probes. The probes were based on software from AbleLink, but each had a number of features and design elements that were implemented during the probe study. Some of the added features were unique to each user, based on her abilities and needs (such as the Memo Recorder and Bus Arrival Information Application), and other features were appropriate for both users (such as the screen lock). I complete the chapter with a summary of the usability requirements identified through the probe study. In the remaining chapters, I will discuss how the usage of the probe informed our understanding of requirements at the level of the sociotechnical environment, and for the design of a meta-design environment.
117
Hey Melissa, what’s up? Love ya, bye! – Voicemail message to me from Linsey
8
Remote Communication Patterns of Two Families 8.1. Introduction The technology probe study illuminated aspects of two families’ socio-technical environment, which included caregivers, care recipients, and technology. The combination of quantitative and qualitative data provided a multi-faceted view of remote communication in the context of the family system. Each family perceived, incorporated, and modified the technology probe differently, reflecting the distinct characteristics of each family system. In this chapter I describe the usage of the probe and the role that it played in each family’s larger goals of safety, independence, and social connectedness. This analysis serves two purposes. First, it gives us a view of the patterns of remote communication between caregivers and care recipient, which is an integral part of the larger family system. Second, these patterns of interaction inform our understanding of how two families with young adults with cognitive disabilities conceptualize and address their goals of independence, safety, and social connection. Both of these aspects of the socio-technical environment have important AT design implications.
118
Throughout this chapter I refer to the technology probe simply as a phone, which was its primary function, and how it was perceived and referred to by the families. 8.2. Linsey and the Roberts Family: Mobile Phone as Connector In this section I will describe patterns of Linsey’s phone use in the context of her daily activities. Through the study it became clear that the phone was an important social connector between Linsey, her parents and her live-in companion Abby. Largely by her mother’s design and support, Linsey is a very active young woman with various vocational and recreational activities throughout her day. As described in Chapter 6, Linsey lived in an apartment approximately one mile from her parents’ house. She stayed in her apartment with Abby during the weekdays and at her parents’ house on the weekends. Linsey’s parents and Abby worked full-time during the day. When Linsey was not involved in an activity, she was usually home alone in her apartment watching TV, napping, or talking on the phone. 8.2.1. Patterns of Remote Communication and Memo Usage The data collected during the technology probe study show that Linsey used her phone intensively and made many calls to her parents each day. Figure 8.1 shows a graph of the number of calls she made over a typical seven day period during the study. This graph shows that Linsey called her parents (combining their mobile phones, landline, and mother’s work phone) an average of 16 times a day, and called Abby 2 – 3 times per day.
119
Figure 8.1: Linsey’s number of calls and their recipients over a 7 day period Linsey also received calls on her phone, but she received far fewer than she initiated. In a typical week, her mother called her three to four times, and Abby called once or twice. When I asked Abby and her mother Carin about why Linsey called so much more often than she received calls, they explained that Linsey called so frequently that they rarely needed to call her. Linsey had different calling patterns on different days of the week. Figure 8.2 shows the number of calls Linsey made at different times of day on different days of the week. This figure only shows data for Monday through Saturday, because on Sundays Linsey’s family tended to do activities together and Linsey made very few calls.
120
Figure 8.2: Linsey’s call frequency across the day and across days of the week Who Linsey called at different times of the day was not random, and illuminated how the phone played a role in the larger socio-technical context. Temporally, Linsey’s calls can be grouped into the following categories: 1. Morning calls: Linsey usually called her mother and her father when she woke up in the morning. The early morning calls (before 7 AM) were always to her father, who woke up very early to commute to work. 2. Daytime calls: During the day Linsey called the broadest range of people, including members of her extended family. Her calls increased when an activity was approaching and just after an activity ended. These calls were usually to her parents and Abby. 3. Evening calls: Linsey usually made a few evening calls to her parents’ house on weeknights. The timing of Linsey’s calls corresponded with her activities, which varied on different days during the week. Table 8.1 represents a calendar of Linsey’s weekly activities, showing her vocational, recreational, and learning activities. Linsey had multiple methods of transportation to her activities: she took the bus independently to the recreation center and retirement home, received private bus transportation to her 121
job at the restaurant and Day Program on Mondays, and got a ride from her parents or Abby to all other activities. Monday
Tuesday
Wednesday
Thursday
Friday
Restaurant: Job (sorted and rolled silverware in napkins)
Restaurant : Job
Saturday
7am 8am 9am 10am
Community Center: Day Program (recreation/ life skills activity)
11am
Community Center: Journal Class
Restaurant: Job
12pm 1pm 2pm 3pm
Rec Center: Spinning
4pm
5pm 6pm
Retirement Home: Job (gave water to residents)
Rec Center: Aerobics Health Club: Daycare Job
Rec Center: Walking
Health Club: Daycare Job
7pm 8pm
Rec Center: Volleyball
Home: Mtg with Melissa
Table 8.1: Linsey's weekly schedule of activities (Monday – Saturday) Information about Linsey’s daily activities – which did not always follow her planned schedule – was also provided by her caregivers’ evening diary voicemail messages and memos. Linsey’s phone had a memo feature that provided Carin or Abby the ability to record a single message, and Linsey could listen to it throughout the day. The memo messages commonly described Linsey’s activities for the day and who was responsible for providing for her transportation. All of the memo messages were downloaded, transcribed and included in the data analysis. Here is an example of a typical journal entry recorded by Carin:
122
Good morning, Linsey, today is Tuesday, October 17. Today you will go to the ACL for your journal class at 10 o'clock. Abby will take you, and Abby will pick you up at 11:30. I have to check with Karen on status of arts and crafts, and if we can make it happen we will. And arts and crafts goes from 3:30 until 5. I'll have Dad pick you up, and get you to work, and then you'll come home and see Melissa. Now that's for a full day! Don't forget … watch what you eat! Love you! From this memo, we get a sense of Linsey’s high level of activity (and the amount of coordination that her mother put into each day!). Linsey’s caregivers recorded a new daily memo for her approximately 60% of the time. On days when Linsey had a new memo, she listened to it multiple times during the day. She typically listened to the memo three to four times soon after she woke up and occasionally between calls. The probe data provided evidence that Linsey enjoyed the memos (in that she consistently listened to them), but there was no correlation between her listening to memos and calling patterns. That is, Linsey’s number of calls did not increase or decrease when she had a memo recorded for that day. Viewing Linsey’s calling patterns in combination with her activity schedule, it is clear that many calls took place right before and right after activities. For example, Linsey worked at the restaurant from approximately 9 AM to 12 PM on Saturdays (in reality this varied somewhat, the driver sometimes picked her up late, and dropped her at home early). Her peak calls on Saturday were from 8 – 9 AM and 12 – 1 PM. Similarly, on Mondays Linsey had an all day recreational activity from 8 AM to 4 PM, and then a water aerobics class from 5 – 6 PM. Linsey made the most calls on Monday between 7 – 8 AM and 4 – 5 PM. Most of the calls before and after activities were to her parents and Abby. In order to understand the nature of the calls (the purpose of the calls, and what was communicated), We turn to data collected through the phone call journal activity. 8.2.2. Linsey’s Goals and her Use of Remote Communication Linsey was not able to articulate how or why she used the phone. To explore the nature of her phone calls, Linsey’s parents and Abby kept a phone call journal for one week towards the end of the probe study. First I met with Carin, Linsey’s father Dan, and Abby and asked them to recall recent calls with Linsey and put them into categories based on purpose. Collaboratively, we developed eight categories of 123
reasons Linsey called. Then the caregivers each kept a journal of received and sent phone calls with Linsey (see Chapter 6 for a picture of a journal entry), and assigned one or more categories to each call. The journals recorded 38 calls over 6 days. Of the journal entries, 74% were categorized, and of the categorized entries 82% were in more than one category. Table 8.2 presents the categories along with the percentage of calls in each category, and an example from a journal entry. Category (Purpose of Call)
Percentage in this Category
Description and Example
Reassurance
64%
Abby: Linsey calls me in the morning to talk about what’s going on this week. She has her schedule memorized; she just wants reassurance about what’s going on.
Time (asking what time it is)
50%
Carin: Linsey hates to be late, and being that she cannot tell time she calls numerous times asking what time it is – for example, she called me many times asking where I was so she would not be late to her 2pm dentist appointment.
Coordinating (current activity)
43%
Carin: Linsey called to ask if she should leave for the bus stop now to go to spinning class.
Boredom
29%
Abby: Linsey left a message on my voicemail, letting me know that she had fun at her journal class and played with computers.
Social Connection
25%
Dan: Linsey calls first thing in the morning, while she’s still laying in bed watching TV. She just calls because she knows I’m on the road. She asks how I’m doing, where I am going. She’s just touching base.
Planning (future activity)
14%
Carin: Linsey called because she was thinking about the overnight party she was attending that night through the recreation center and was worried about spending the night.
Reminding (reminding her/reminding me)
18%
Abby: Linsey called in the morning to remind me that I was taking her to work at the retirement home.
Security/Safety
7%
Abby: Linsey left a voicemail message to tell me that Mom was not at the apartment yet (she was supposed to be there).
Table 8.2: Purpose of Linsey's phone calls: categories and examples From the table, we see that reassurance was the most commonly applied category in describing Linsey’s calls. Linsey sought reassurance throughout her day for various reasons: waiting at the bus stop, waiting at home for someone to pick her up, or
124
feeling anxious about a particular upcoming activity (like going to the doctor for a flu shot). These calls tended to be short, but she would often call back repeatedly within a short time span. Reassurance frequently co-occurred with other categories, especially time and coordinating. As the examples illustrate, most of Linsey’s calls had an informational purpose (such as asking where someone is) and an emotional purpose (seeking reassurance that she was not going to be late, or that she was not forgotten). The underlying emotional motivation of Linsey’s calls became acutely evident at one point in the study. After the phone call journal activity, time emerged as a major reason for Linsey’s calls: she was always calling her mom, dad, and Abby to ask what time it was. In the subsequent participatory design meeting, the family and I decided to add a visual and audio display of the date and time to Linsey’s phone (described in detail in Chapter 7). With this feature, Linsey could push a button and the phone would announce the time, e.g., “It is ten fifty seven AM.” According to the hypothesis that the purpose of Linsey’s calls was to find out the time, this should have dramatically reduced the number of Linsey’s calls. In fact, this did not change Linsey’s calling patterns, and she rarely used the time feature (although she demonstrated that she could use it). This is one example of how the evolving nature of the probe allowed us (the family and me) to formulate, explore, and evaluate hypotheses. In this case, we gained support for the hypothesis that there was another motivation for Linsey’s informational requests, and it seemed to be emotional reassurance and desire for a social connection. The emotional motivation of Linsey’s calls appeared with both positive and negative emotions. Linsey initiated calls when she was feeling anxious, nervous, or upset (such as when her parents were running late picking her up). She also initiated calls when she was feeling excited, joyful, or looking forward to something. The phone call journals included a number of reports where Linsey called to share “good news,” such as her mother’s out-of-town friend coming to visit. The high number of calls made on Friday afternoon usually corresponded to a special activity happening that evening, such as a dance, sleep-over, or a social event with friends. There was a very predictable spike in phone call activity on Friday afternoons between 4 PM and
125
6 PM. Linsey’s parents and Abby reported that Linsey called repeatedly to share her excitement about the upcoming event. Another time period when Linsey made many calls – often one right after the other – was when she woke up in the morning alone in her apartment. At this time she tended to call her dad who was commuting to work, or her aunt Dorothy, with whom she held relatively long conversations. When she did not reach one family member, she called another. When she connected with someone, she talked about a variety of topics, including: her dreams, the weather, and upcoming events. It appears that she was seeking an emotional connection with others during these calls when she was home alone. The data provides strong evidence that Linsey’s phone played an important role in social and emotional connection with her parents and Abby. Linsey is a social person: she wants to know about and be involved in everything that’s going on in her parents’ lives, her extended family and friends’ lives, even her parents’ friends’ lives. Linsey is also a very emotional person. She becomes excited quickly and likewise becomes anxious and upset quickly. 8.2.3. Caregiver Goals for Remote Communication Linsey’s main goal with the phone was to be socially and emotionally connected with her caregivers. Linsey’s caregivers had broader goals for her remote communication system that fit into their larger goals for their daughter concerning safety, social connection, and independence. 8.2.3.1. Safety Goals Like all parents I interviewed, Linsey’s parents held their daughter’s safety as their top priority. The most important safety goal was that Linsey have her phone with her at all times, and that it be charged and functional. Linsey’s parents and Abby reinforced behavior to support these goals, such as making sure Linsey charged her phone every night. Linsey became quite attached to the phone during the study – and in general carefully guarded her belongings – so ensuring that she had her phone with her was rarely a problem. To address the goal that the phone be functional, the
126
software needed to be stable and well tested. The caregivers needed to know that I was available to provide technology support quickly if something went wrong with the phone. We also added a way for Linsey to start up the phone herself if the software crashed or the battery ran out, to reduce her dependency on others to recover from an error (this is described as the “Accessible Launcher” in Chapter 7). Knowing that Linsey had her phone and could call many people in her caregiver network was a key benefit of the phone from the caregivers’ perspective. Carin explained that Linsey called so frequently that sometimes Carin could not answer the phone while she was at work, but she did not worry because she knew Linsey could call others if something was wrong. Having the accessible contact list enabled Linsey’s mother to be more reliant on Linsey’s network of caregivers. 8.2.3.2. Social Connection Goals Unlike other young adults with cognitive disabilities in Phases 2 and 3, Linsey had a strong desire to be socially connected. Making sure Linsey stayed in contact with her caregivers was not a problem – in fact, Linsey’s mother was constantly trying to teach Linsey to call her less often. Carin attributed her daughter’s frequent calling to the fact that she loves to chat: She loves to talk on the phone. And it’s always been a concern, it’s something we’re trying to work with. She’s very good at redialing. And uh, doing it numerous times over and over, and focusing on helping her to better understand that she can’t be calling me at work, or someone repeatedly, and when someone says goodbye, she needs to understand that it’s time to say goodbye and hang up. Or she’ll kind of, she likes to continue to talk about the day, the weather, activities, whatever. But she’s a little chatterbox. In interviews, Carin never directly associated Linsey’s high calling frequency to her need for emotional support and connection. Instead, she repeatedly explained that Linsey loves to chat, and that she was constantly encouraging Linsey to chat less (especially with Mom and Dad when they were at work). One interpretation of Carin’s characterization is that she was consciously or unconsciously simplifying Linsey’s calling behavior to “liking to chat” as a coping strategy to approach the problem. Characterizing Linsey’s behavior as “liking to chat” simplified the goal of
127
modifying this behavior to teaching her daughter to chat less often. This had fewer implications and challenges than teaching her daughter to be less emotionally reliant or meeting her emotional needs in other ways. This is an important point because it illustrates how caregivers’ perspectives and descriptions are inherently tied to their own goals, needs, and coping strategies. In addition to limiting Linsey’s “chatting,” Linsey’s parents’ goals around social connection included broadening Linsey’s social network. When we added family members to Linsey’s contact list, Carin was proud that her daughter gained the ability to keep in touch with extended family. Carin reported that Linsey called her relatives when her niece had her first birthday, and sang happy birthday “all on her own.” Linsey’s parents wanted her to have a circle of friends locally who would provide regular social activity. In spite of Carin’s effort and Linsey’s social nature, Linsey had very few friends her age with whom she maintained regular contact. Carin attributed this to a lack of effort of other families to maintain social relationships. Linsey also demonstrated selective social interest: she did not call everyone in her contact list. She preferred to call certain people (mainly her parents, Abby, and some extended family) and rarely called two local friends her age whose contacts were added by Carin. 8.2.3.3. Independence Goals Linsey’s parents and Abby were constantly working on increasing her independence and teaching her important life skills. These did not all revolve around remote communication. Carin would schedule training classes for Linsey in various life skills, such as cooking, learning more bus routes, and keeping physically fit. Abby and Linsey would regularly plan menus focused on healthy eating habits, and Abby would coach Linsey to pick out healthy food at the grocery store. Her caregivers encouraged Linsey to take initiative and do things on her own. Remote communication technology enabled Linsey’s caregivers to provide support when Linsey was doing new activities and she did not exactly know the routine, or when the situation strayed from the plan. For example, Linsey occasionally took the bus by herself to the neighborhood grocery store. Because this
128
was not part of her regular routine, Linsey would call Abby and her parents a few times along the trip for help with planning and coaching. For example, she would ask Abby when she should leave the house to catch the bus, and she would ask her mother if she had enough money to buy certain items. For unfamiliar or less structured activities, the phone played a key role in supporting Linsey’s independence. Carin hoped that the remote communication could act as scaffolding to support Linsey as she learned to be independent, and that Linsey would need to make fewer calls to her caregiver network over time. One of Carin’s primary independence goals for Linsey was that she gain the ability to make more of her own decisions without checking in with her caregivers. There is a clear tension between Carin’s goal for Linsey’s independence and Linsey’s goal for a close social connection. This is an important example of how caregiver and care recipient goals are not always aligned. 8.3. Margaret and the Smith Family: Mobile Phone as Emancipator We will now turn to the second family in the probe study, Margaret Smith and her parents Sally and John. The Smith family’s perception, usage, and incorporation of the probe into their socio-technical environment were very different from the Roberts family. For Margaret the probe was primarily an emancipator: it gave her functional support and gave her parents more peace of mind to enable Margaret to achieve her desire to stay in her apartment independently. As described in Chapter 6, Margaret’s abilities and needs were significantly different from Linsey’s. In disability parlance, Margaret would be called “higher functioning” than Linsey – she could read and write at a first grade level, communicate with a larger vocabulary, and tell time (although she had trouble reasoning about time). Linsey and Margaret’s functional abilities are described in more detail at the beginning of Chapter 6. Before the probe study, Margaret frequently used a mobile phone to keep in contact with her parents. She had memorized a handful of numbers, including her parents’ landline, her dad’s office number, and each of their mobile phone numbers.
129
At the beginning of the probe study Margaret had recently moved into her own apartment, although she usually stayed the night at her parents’ house. Margaret felt ready to move out of her parents’ house, but her parents Sally and John had safety and social isolation concerns about Margaret staying there alone. During the months of the probe study (for reasons related and unrelated to the study), Margaret increasingly spent more time at her apartment alone. By the end of the study, Margaret stayed the night at her apartment two to three times a week. Because Margaret was already capable and in the habit of calling her parents from multiple phones, the technology probe was not Margaret’s primary means of communicating remotely with her parents. Due to the research nature of the probe, both Linsey and Margaret carried a secondary mobile phone (their previous mobile phone) in case of emergencies. This secondary mobile phone was connected to their old phone number, and the probe phone had a new number. During the probe study, Margaret continued to use her previous mobile phone (as well as her landline phone) to communicate with her parents. In particular, Margaret and her mother Sally had developed a deep routine of calling each other on their mobile phones and this proved a very hard habit to break – not only for Margaret, but even more so for Sally! Throughout the study, Sally primarily called Margaret on her previous mobile phone number. This became evident during the phone call journal activity, when Sally’s reported calls were far greater than the calls logged on the technology probe. Sally explained that this was simply out of ease and habit – although she put Margaret’s new phone number into her mobile phone, Margaret’s speed dial number still went to her old mobile phone number. She often called Margaret when she was in the car or multi-tasking and so she almost always used the speed dial number. At each meeting Sally would comment about needing to call Margaret more on her new phone. Unfortunately the speed dial explanation did not come up until the end of the study, at which point it was too late to help Sally change her speed dial number to Margaret’s new number. Since Sally and Margaret were communicating mainly through Margaret’s old phone, the probe data did not give a complete picture of Margaret’s remote communication patterns with her parents like it did Linsey’s. The phone call journal
130
data was more informative for Margaret’s analysis in that regard. The probe data did reveal other patterns in Margaret’s communication behavior. 8.3.1. Patterns of Remote Communication and Technology-Supported Tasks With her previous mobile phone, Margaret could only call contacts whose phone number she had memorized. With her new phone, Margaret had a picture-based contact list that she used to make calls. Margaret began communicating with new contacts, including her sister, her brother, her aunt and uncle, her care companion Donna, and her bus driver. Table 8.3 summarizes Margaret’s remote communication contacts before and during the study. Margaret’s Call Recipients with Previous Mobile Phone
Margaret’s Call Recipients with Probe Phone
• Parents’ house • Mom’s mobile phone • Dad’s mobile phone • Dad’s office • 911 (if in an emergency)
• Parents’ house • Mom’s mobile phone • Dad’s mobile phone • Dad’s office • Jane (sister) • Tom (brother) • Aunt and Uncle • Bus Driver (CallnRide) • 911
Table 8.3: Margaret's call recipients before and during the probe study In general, Margaret did not make many calls. This was true before and during the probe study. Sally and John explained that Margaret did not have much interest in making social calls (as described in Chapter 5) and mainly used the phone to coordinate with her parents. Figure 8.3 shows a graph of the total number of Margaret’s calls over the entire probe study (unlike Linsey’s chart, which shows the weekly average).
131
Figure 8.3: Margaret's number of calls and call recipients over total 8 weeks Margaret made the most calls (35) to the probe study voicemail diary, which was a “required” call during the study. Margaret’s father John required that she keep her own voicemail diary, rather than having the parents record it each night (which had interesting implications for the data collected). John felt that this would be a way for her to have a responsibility during the study, as well as practice using the phone on a regular basis. During the probe study Margaret made occasional social calls to her sister, brother, uncle, friend Andrea and care companion Donna. This was new behavior, and indicates that the probe broadened Margaret’s social calling circle as it did Linsey’s. Margaret called me on days when I was coming over to visit, to confirm our meetings. Margaret also called her Call-n-Ride bus driver Rita (Call-n-Ride is a bus service for people with special needs) while she was waiting at the bus stop to find out where she was. Margaret usually told her to hurry because it was cold (the study took place during a cold Colorado winter)! In addition to supporting mobile phone communication, Margaret’s probe had features that might be better described as facilitating technology-supported tasks rather than remote communication tasks (although the supported tasks were previously accomplished through remote communication). John and Sally wanted the probe functionality to replace some of their calls with Margaret. In other words, they wanted to transform some of their remote communication tasks through the technology, so that the call was not needed at all. The suggested task transformations
132
came up during an early design activity when we were analyzing Margaret’s remote communication tasks, and sketching paper mock-ups of technology to support them. Figure 8.4 describes the two tasks that were transformed from phone communication to probe features. Margaret rode to work part of the way on the Calln-Ride bus, and part of the way on a public bus. While waiting at the bus stop for the public bus, Margaret often called home if the bus was a few minutes late. Margaret’s mother usually told her to wait 5 more minutes and call again if the bus had not shown up. Margaret’s parents envisioned that if Margaret’s phone could somehow tell her when the bus would arrive, she would not need to make those phone calls. The second task was Margaret’s wake-up call. Margaret had to wake up at a different time each day. Margaret’s parents had trouble remembering when she needed to wake up and sometimes forgot to wake her altogether. They felt that technology could take over this burden and would do a better job than they could! These two probe features are described in detail in Chapter 7. In the new tasks, the technology directly provides the support that previously required communication with a parental caregiver. The ability to reframe some of Margaret’s tasks from remote communication tasks to technology-supported tasks was an affordance of the participatory and taskcentered nature of the research methodology. Focusing on understanding and supporting the family’s remote communication tasks, rather than starting with an a priori conception of the probe-as-a-mobile-phone, allowed the family’s goals to be reframed as features that did not actually require phone calls.
133
Transforming Two Remote Communication Tasks
Task 1: Margaret calls home because the bus is late Margaret is at the bus stop on Main Street, coming home after work. The bus is late, so she calls her mom’s cell phone to ask for a ride. Mom tells her to wait 5 more minutes, and then call her back if the bus still hasn’t arrived, and she will come pick her up. Transforming the Task from Remote Communication to Technology-Supported New Task: Margaret checks the bus arrival time because the bus is late Margaret is at the bus stop and the bus is late. She takes out her phone and navigates to the bus arrival information screen. The screen tells her (in real time) how many minutes she will have to wait at the bus stop for the bus to arrive. She can decide to call her Mom for a ride if the wait is too long.
Task 2: Margaret’s parents remind her to get up and prepare for work Each morning, Margaret’s schedule is different. When she stays in her apartment, her parents call her to wake her up and tell her to start getting ready for her activities. On Tuesdays and Thursdays, Margaret’s parents wake her up at 9 AM and remind her that her bus will pick her up to go to work at 10:30 AM. Transforming Task from Remote Communication to Technology-Supported New Task: Margaret’s phone wakes her up and reminds her to get ready for work Each morning, Margaret’s phone alarm goes off at the right time to wake her up in time for that day’s activities. The phone has a recorded reminder for that day, e.g., “Good morning. Today the bus will pick you up at 10:30 for work.”
Figure 8.4: Two remote communication tasks that were transformed into technology-supported tasks Understanding Margaret’s usage of the probe includes observing her patterns of using the bus arrival information screen, acknowledging reminder alarms, and sending and receiving phone calls. Figure 8.5 demonstrates how frequently each of these three features was used throughout the study. Note that between December 19
134
and January 2 there was a break in the study when the family went out of the country on holiday.
Figure 8.5: Margaret's pattern of probe usage across the study period The most consistently used feature of the phone was the bus arrival information screen. Margaret rode the bus to and from work on Tuesdays and Thursdays. Each day that she waited at the bus stop on her way to work, she pressed the “Work” button around 10-15 times, once every 30 seconds or so, to update the display that told her how many minutes longer she needed to wait. While she could not explain to me why she pressed the button this many times, it indicated that she may not have a good sense of time periods (e.g., how long a minute was) – this explanation was supported by her parents’ description of her limited understanding of time. From her behavior, it is clear that Margaret understood that pressing the button would update the screen and that she wanted to be frequently updated about the location of the bus. This functionality reduced the number of calls to her parents when the bus was late, most likely because having the information available to her gave her an increased sense of control and reduced her anxiety. Margaret’s reminder alarms were not as highly used as the bus arrival information. Margaret had a morning alarm on Mondays, Tuesdays and Thursdays (when she had morning activities, shown in Table 8.4), and a nightly alarm that went
135
off every night to remind her to call the diary. Figure 8.5 illustrates that during most of the study, Margaret only acknowledged her alarms once or twice a week. Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
7am 8am 9am 10am 11am
Community Center: Day Program (recreation/ life skills activity) (The same activity as Linsey)
Special Olympics ski team training Work: Grocery store
Work: Grocery store
Life skills training with Donna
Life skills training with Donna
12pm 1pm 2pm 3pm 4pm 5pm 6pm 7pm 8pm
Table 8.4: Margaret's weekly schedule of activities There were several reasons why Margaret did not acknowledge the reminder alarms very often. A primary reason, surprisingly, had to do with the phone’s charger. Margaret’s parents had a deep routine of keeping Margaret’s phone charger plugged in downstairs near the front door. By keeping it there, they could confirm that she remembered to plug it in at night and make sure she would not forget to take it in the morning. However, Margaret’s bedroom was upstairs – so when the morning alarms went off, the phone was downstairs and she did not hear it. Sometimes Margaret’s parents would hear it and it would remind them to go upstairs and wake her up (which was a different but still effective way to incorporate the technology into their system). When Margaret stayed at her apartment she did not take the charger, so the phone 136
would often go dead (the battery lasted less than 24 hours) and the alarms would not go off. After observing this pattern for several weeks, I brought the family two more chargers: one for Margaret’s bedroom and one for Margaret’s apartment. After that Margaret began to acknowledge her alarms more regularly. This is one example of how technology peripherals (like a charger) can have a big influence on how technology is adopted and used. These factors must be considered as an important part of the design. Another reason why Margaret may have not acknowledged her reminder alarms consistently was that she did not seem to like them. She never offered her opinion of the alarms during the interviews or design meetings (perhaps because her parents emphasized that she must use the alarms if she wanted to stay in her apartment by herself). One day when I was visiting Margaret at her apartment and we were looking at the phone, I played one of the morning alarm messages. It said, “Good morning! It’s Tuesday. Today you go to work. You need to be ready and waiting at the stairs at 10:30.” (The voice was mine, and the message was authored by Margaret’s mother.) Margaret and I then had the following dialog: Margaret (irritated): I know that already! (referring to the alarm message) Me (surprised): Oh. You know that already? Do you like to hear the message in the morning? Margaret: No. I already know that. Another example occurred during a morning meeting toward the end of the study, when the family and I were sitting around the kitchen table. The phone was sitting on the table and the wake-up alarm went off. Margaret looked annoyed and said, “Oh god,” and pressed the screen twice quickly: once to stop the alarm, and once to stop the message (without listening to it). I hypothesized that Margaret did not like the alarm because the messages had information that she felt she already knew (she was very good at remembering her schedule) and perhaps she even perceived it as an extension of her parents’ control – all of the alarms had been designed by Margaret’s parents. To explore this further, one day I casually asked Margaret if she would like to add any reminders of her own. She suggested that we add reminders for her two favorite TV shows –both
137
professional wrestling shows. I recorded a message for one show, “It’s time for Friday Night Smackdown!” and Margaret recorded a message for the other one, “Monday Night Raw!” The purpose of this activity was to design alarms that were driven by her desires, not just her parents’, and give Margaret more of a sense of ownership and control of her reminders. A few days later in a meeting with the family Margaret proudly described the new reminders and wanted me to play the recordings for her parents, who were very amused. It seems that this was an effective way to increase Margaret’s perception of the phone as hers and expressive of her interests, not just her parents’. 8.3.2. Margaret’s Goals and Use of Remote Communication To understand the purpose and the content of Margaret’s phone calls with her parents, Margaret’s parents kept a phone call journal for one week during the probe study. Before the journal activity, Margaret’s parents and I discussed examples of calls with Margaret, and collaboratively developed categories for the journal. Safety check-ins and information requests were described as the primary purpose of remote communication. During the journal activity 21 calls were logged over six days. The calls were assigned to one or more categories shown in Table 8.5. Many of the calls were for coordinating activities, and to check-in on Margaret when she was in her apartment alone. For example, Margaret’s parents required that she call them whenever she planned to leave the apartment (e.g., to take out recycling or check the mail). She called again as soon as she returned. Margaret’s mother also called her periodically throughout the day to make sure her door was locked, she was eating, and that everything was okay. The desire for independence was an important motivator in Margaret’s remote communication patterns. Her parents set up structures of communication, such as the calls before and after Margaret left her apartment, that Margaret had to follow in order to stay in her apartment independently. Margaret and her parents had numerous safety check-in calls on days when Margaret was staying at her apartment.
138
Category (Purpose of Call)
Percentage in this Category
Description and Example
Social/ Reassurance
48%
Margaret called Dad at lunchtime from the ski lodge to tell him about her Special Olympics ski practice. She was proud because she had a good ski day.
Planning
33%
Margaret called Mom from the ski bus to tell her that the bus was half an hour from Boulder, so she would need to get picked up soon.
Safety
24%
Margaret called Dad from her apartment to tell him she was going outside to take out the recycling. She called back five minutes later to tell him she was back in the apartment and the door was locked.
Check-In
19%
Mom called Margaret before she went to bed to make sure that Margaret had eaten dinner, and that the door was locked.
Reminder
14%
Margaret called Mom from her apartment to confirm when Mom was picking her up to go to the movies with a friend.
Information Request
10%
Margaret called Dad at lunchtime from the ski lodge to ask, “Where is my lunch money?” He told her it was in the inside pocket of her ski bag.
Table 8.5: Purpose of Margaret's phone calls: categories and examples The most surprising finding was that the most common assigned category was “social/reassurance,” which had been very de-emphasized in all the earlier meetings. In the journals, Sally and John provided various examples of calls that had a significant social/emotional connection component. For example, on nights when Margaret would stay the night at her apartment she would usually call home in the morning just to say “Good morning.” Margaret’s voicemail diary messages also indicated her desire to connect socially. She was instructed to call each evening and describe: 1) any calls today, 2) any new ideas, and 3) any problems. In practice, Margaret reported a few calls she made (although she left out most) and never reported any ideas or problems (even if the phone was having problems). As the study went along and Margaret and I developed a friendly relationship, she would leave social and informational messages in her voicemails. For example, once she left a voicemail in the morning to tell me she had an orthodontist appointment and was a little worried. Here are some other examples:
139
Hi Melissa, this is Margaret. I just wanted to say thank you for coming to my apartment. I really enjoyed having you there. And I went to the movies, and had fun! I saw Night at the Museum. It was fun and funny. So I think I will see you probably the day after tomorrow. Thank you. Bye! Hi Melissa, it’s Margaret. I enjoyed your visit. And I did watch wrestling. I’m at my parents’ house tonight, and I’m going to spend the night at my apartment on Tuesday. I have no questions, I have no ideas, and I’ll talk to you soon. Bye! Hi Melissa, this is Margaret. I’m watching Friday Night Smackdown wrestling tonight, and my sister came by, and I was really happy and really proud. And I’m sorry I missed two calls. It got real late last night, it was about 11 when I went to bed, so I’m calling tonight to let you know. Thank you, bye! I provide these three examples because they each illustrate people and things that appear to be important and exciting for Margaret. First, Margaret was very pleased and proud when someone visited her in her apartment. If her sister or I visited her, she would mention the visit for a few days in the diary voicemails. Second, Margaret talked about her sister Jane frequently in these messages, indicating that Jane played an important role in her life. Third, Margaret often talked about staying the night at her apartment – when she was staying there next, what she would do (e.g., watch wrestling), and when she went to bed when she was staying alone. In general, a central aspect of these calls was to share positive and negative emotions. A third indication that Margaret desired social connection was that she specifically requested to have family members – cousins, and an aunt and uncle – added to her phone contact list. During early design meetings Margaret did not indicate that she had any interest in designing the contact list, but when she saw pictures of her parents and siblings she enthusiastically asked to have more family members added. Finally, at the closing interview of the study I asked the family about perceived benefits and value of the probe. The parents cited their increased confidence in Margaret’s ability to be independent. When I asked Margaret her favorite thing about the new phone, she simply said, “My sister.” After the probe study ended there was further indication that Margaret missed some of the social activity of the study. A few weeks after the end of the probe study, I went to lunch with Margaret. She stated that she would like to start leaving the
140
voicemail diary messages again. I asked her what she missed about them and she explained that she liked talking about her cousins, her sister, and her uncles. 8.3.3. Caregiver Goals with Remote Communication Like Linsey’s parents, John and Sally had goals for Margaret’s remote communication system that fit into their larger goals for their daughter concerning safety, social connection, and independence. 8.3.3.1. Safety Goals John and Sally’s goals for Margaret’s use of her phone largely revolved around safety. In the Phase 1 and Phase 2 interviews, the family described how Margaret’s mobile phone mainly provided the ability to call one another if anything went wrong. Her parents would call Margaret periodically when she was home alone. They described how this safety connection overlapped with issues of social/emotional connectedness and independence. For example, Margaret used to frequently call home from the bus stop when she was learning to ride the bus just to “check-in.” She gradually stopped calling as she gained confidence. 8.3.3.2. Social Connection Goals John and Sally perceived their daughter to be generally uninterested in social connections with others. They were concerned that she would become socially isolated if she stayed in her apartment too often, which was one reason they did not let her stay there all the time. At the beginning of the probe study, they said that Margaret did not have any interest in talking on the phone socially. Their perception changed somewhat during the study, as Margaret made calls to her sister, and requested to have her cousins, aunt and uncle added to her phone contact list (although she rarely called them). During the study, her parents even suggested that the phone could remind Margaret to make calls to friends and loved ones with whom she had not communicated in a while. Yet there also seemed to be an element of parental control to determine when and who Margaret called. Margaret often felt uncertain whether or not she had permission to make the calls. During one meeting she asked her parents, “Can I call Uncle Joe tonight?” Her parents said yes, but her 141
mother reinforced her position of control by adding, “Now don’t start calling him all the time.” It was very evident that Margaret and her parents had a close relationship, and much of their communication was social and emotional in nature. Margaret and her mother were particularly close, and would talk on the phone a few times a day if they were not together. Unlike the situation with Linsey’s caregivers, there was not an issue of Margaret calling too much. On the contrary, Sally’s concern was that Margaret would go to her apartment and forget to call anyone. 8.3.3.3. Independence Goals Margaret’s parents were constantly balancing their concerns regarding her safety and social isolation with their goals of independence for Margaret. During the study, Margaret was transitioning to a much greater level of independence, living in her own home away from her parents. While Margaret expressed that she was ready for this transition, her parents’ approach was more cautioned and incremental. They maintained quite a bit of control over Margaret in her apartment. A safety-motivated example is their requirement that she call before and after leaving the apartment. They expressed their concern about social isolation by installing a television monitoring device that only allowed Margaret to watch a certain number of hours of television per day. They felt that otherwise Margaret would watch too much TV and never want to leave the apartment to see friends and family. Margaret’s parents viewed some of the probe features, particularly the bus arrival screen and reminder alarms, as important new ways for Margaret to gain independence. John was very proud that Margaret used the bus arrival information screen and no longer called home every time the bus was late. Sally explained that knowing Margaret had this information available to her was a “comfort” for Sally, because it gave her daughter more self-reliance. While the reminder alarms were not used as successfully, John and Sally perceived this feature as an important support for Margaret to become more capable of living independently.
142
8.4. Summary This study provided an in-depth understanding of the nature of remote communication between two families, and the role that it played in those families’ larger goals of independence, social connection, and safety for their daughters. When presenting an analysis based on empirical data, it is important to keep in mind that different research methods will inform our understanding of phenomena in different ways. A survey or broad interview study with many participants can identify important patterns, but usually will not give us a strong understanding of why things are happening – the motivations of the participants – or the context in which they take place. Conversely, an in-depth case study with two families is not an appropriate method for making generalizations about the prevalence of phenomena in the entire target population. However, this method is very useful for providing deeper insights and details to illuminate patterns observed in a larger population [Kuniavsky, 2003]. In this chapter I described each family’s use of the technology probe in detail. In Chapter 9 I will draw from the findings across all three phases of my research as well as related studies to discuss AT design implications.
143
Family communication patterns emerge from the process by which families create and share their social realities. That is, they are inextricably linked to the most basic social functioning of the family. – Koerner & FitzPatrick, 2004, p. 181
9 Social and Emotional Aspects of Assistive Technology 9.1. Introduction The technology probe study gave us a rich understanding of the socio-technical environment of two families in our population. The data gave a glimpse at the complex social system that underlies family communication patterns, and the nuanced tension between parents’ and children’s desires towards independence, social connection, and safety. In this chapter I synthesize my findings with related research to articulate general implications for designing assistive technology for individuals with cognitive disabilities. I turn to literature in the social sciences, particularly communication and social psychology, to gain insight on how to understand the emotional and social aspects of family communication. I present four high-level categories of remote communication in the family, each having informational and relational motives. I 144
discuss the complexity added by motives and priorities that differ between parents and children. These emotional and social issues must be important considerations when designing AT for this population. 9.2. Informational and Relational Aspects of Family Communication In communication and social psychology literature a foundational assumption is that all human communication has an informational aspect and a relational component [Rogers & Escudero, 2004]. When people communicate, they exchange information (the content in the communication) and they interact at a relational level: creating, sustaining, or altering relationships with one another. It is the relational level of communication at which we convey affection or disgust, gain or lose status and power, and construct our social reality. In other words, “We do not relate and then talk, but we relate in talk” [Duncan, 1967, p. 249]. The relational component of communication is always present, whether we interact with strangers or loved ones, and whether or not we choose to acknowledge it. Relational communication is more central in close relationships such as among family members. In close relationships, the individuals in the relationship have a high level of interdependence and a strong and frequent impact on one another. The family, which has been called the archetype of a close relationship [Berscheid & Peplau, 1983], manifests most of the complex issues that occur in close relationships. For this reason parent-child, parent-parent and sibling relationships are a common focus in relational communication research. There are entire journals dedicated specifically to this study, such as the Journal of Family Communication and Family Process. It is no bold statement, and most of us know from personal experience, that emotion is a central component of family interaction. Fitness and Duffield describe emotion as “the currency of family relationships, imbuing them with meaning and importance” [2004, p. 473]. Parent-child relationships in particular are characterized by a high level of “socio-emotional” content [Kelley, Berscheid et al., 1983; Ridgeway & Johnson, 1990]. While each family has their own unique deeply established communication practices, there are common patterns. Studies of parent-
145
child communication have found that parents and children have predictable socioemotional motives when they communicate with each other, including pleasure, affection, inclusion (overcoming loneliness or isolation), and control [Barbato, Graham et al., 2003]. While most communication research focuses on face-to-face interaction, studies of family communication over mobile phones have also shown a strong component of emotional connection and emotional maintenance [Ling, 2006; Palen & Hughes, 2006]. For example, Palen and Hughes refer to the mobile phone’s role as a “tool for remote mothering” [2006, p. 2]. They found that while the nature of remote communication differed with the age of the child, it consistently served to maintain a direct emotional connection between mother and child. In light of the communication theory and empirical studies described above, it is not surprising that remote communication between parents and children observed in Phases 2 and 3 were found to be simultaneously informational and relational in purpose. In Table 9.1, I identify four categories of remote communication and describe their informational and socio-emotional goals. These categories are grounded in data collected in the Phase 2 and Phase 3 studies and are confirmed by other remote communication research [Ling, 2006; Palen & Hughes, 2006]. Still, this categorization system is an early conception that should be empirically evaluated further with other studies and a larger sample size. The four categories of remote communication are conceptual groupings of the categories discussed in the Phase 2 interviews, as well as the categories developed by the Roberts and Smith families during the phone call journal activity.
146
Category
Informational Purpose
Socio-emotional Purpose
Safety check-in
Parents:
Parents:
• Knowing child’s location • Knowing child’s environment is safe
• Emotional comfort and reassurance
(doors are locked, etc.)
Children: • Knowing parent’s location Request for help, getting “un-stuck”
through knowing child is safe
• Emotional closeness with child Children:
Children:
• Emotional closeness with parents • Reassurance Parents:
• Instructive assistance from parents
• Maintaining emotional balance with child
Children: • Emotional support, reassurance, easing anxiousness or distress
Coordinating/ planning/ reminding
Parents & Children:
Parents:
• Exchanging information about the
• Reassurance through knowing that
time, location and other organizational details of upcoming activities
• Agreeing on responsibilities (e.g., parent will pick up child from work)
child is safe and in expected location
Children: • Sense of control in their world through knowing when things are happening, where parents are
• Perhaps emotional support, easing anxiousness
Social/ “Chatting”
Parents & Children:
Parents:
• Sharing details about the day • Exchanging stories
• Emotional closeness with child, maintaining emotional balance with child
Children: • Emotional closeness with parents through emotional co-incidence
Table 9.1: Four categories of remote communication between parents and their children, and the informational and socio-emotional purposes of each category For each category of remote communication, there are informational and socioemotional motives for both the parents and the child. Safety check-ins are associated with parents wanting reassurance that their child is safe as well as an emotional connection/closeness with their child (as parents described in the phone call journal). Similarly children initiated safety check-ins to establish emotional closeness with their parent. When their child requested help getting “un-stuck,” parents not only provided instructive guidance, but also comforted their child and maintained emotional balance 147
[Ling, 2006]. Emotional balance relates to the process of maintaining social order and managing stress in a relationship. Ordinary stressors and upsets can be more traumatic for an individual with cognitive disabilities because of their lack of recovery strategies; parental reassurance can help the individual restore emotional stability. In stressful situations children sought emotional support, which refers specifically to support efforts directed at reducing negative emotions including stress, sadness, fear and anxiety. Parents are a child’s primary source of emotional support as they are growing up, usually until adolescence in US culture [Burleson & Kunkel, 2002]. For families who have children with cognitive disabilities parents may remain the primary source of emotional support for much longer, as was true with the Roberts and Smith families. Communication for coordinating, planning and reminding also served an important emotional purpose, perhaps more for the children than the parents. Through these calls, parents gained reassurance that their child was safe and in an expected location. Children received reassurance and gained an increased sense of control of their world, through confirmation that things were going according to plan. For example, both Linsey and Margaret strongly disliked being late and called their parents when they were late to pick them up. Linsey frequently reminded her caregivers of upcoming activities when she would need a ride. Finally, chatting or purely social calls had the emotional motivation of establishing emotional closeness between the parent and child. One way that family members sustain close relationships is by developing emotional co-incidence, the process of creating a shared positive emotional experience [Fitness & Duffield, 2004]. To create emotional co-incidence, a child might share a proud or exciting moment so that the parent feels the same emotion. Social support literature indicates that when an individual is in a stressful situation, chatting about unrelated topics reduces the stress level [Tardy, 1994]. Finally, chatting gave parents the ability to check in on their children and maintain emotional balance if they were overly excited. Parent-child communication is loaded with emotional content, whether the communication relates to checking in, giving help, coordinating, or just chatting. Another socio-emotional aspect of the parent-child relationship is that the goals and
148
priorities between parents and children are not always perfectly aligned. In particular, this research uncovered tensions between goals related to safety, independence, and social connection between parents and children. 9.3. Relational Aspects of Safety, Independence, and Social Inclusion All three phases of this research identified independence, social connection, and safety as important goals for families with a child with cognitive disabilities. Research has shown that these issues are not unique to these families, and are in fact common goals in general in remote communication among family members [Ling & Yttri, 2002; Palen & Hughes, 2006]. In this section, I discuss how the technology probe study has enriched our understanding of the complexity of these three concepts and their role in remote communication. Specifically, we find two types of tensions: first, there is tension between the three goals of safety, independence, and social inclusion; second, there is tension between the goals and priorities of caregivers and care recipients. Safety, independence, and social inclusion are not unrelated, but rather have a complex interconnected relationship (shown in Figure 9.1). Establishing a basic level of safety is a pre-requisite for independence and social connection. Having a strong social network can play a role in monitoring one’s safety. There are also conflicting objectives between goals for safety and independence, and between goals for safety and social inclusion. A child’s independence or autonomy implies less intense parental monitoring and control, and more empowerment to make decisions. Independence combined with societal inclusion implies being “out in the world” in a fairly uncontrolled environment, where there are more opportunities but greater risks. The tension between security and an independent, adventure-filled life is described in a classic quote by Helen Keller [1957], where she argues for adventure over safety: Security is mostly a superstition. It does not exist in nature, nor do the children of men as a whole experience it. Avoiding danger is no safer in the long run than outright exposure. Life is either a daring adventure or nothing. In my research studies, I found that parents were constantly balancing their objectives for their child’s safety, independence, and social connectedness. They
149
valued remote communication for giving them the ability to remotely monitor and stay “constantly connected” with their children as the children participated in community activities independently.
Figure 9.1: Diagram of the interconnections and potential tensions between safety, independence, and social connectedness A second tension exists between the goals and priorities of parents and children. The issue of a child’s independence versus safety and parental control is a common source of conflict in adolescence among typical families in our culture. During this transitional period, the eventual transfer of control from parent to child is generally seen as inevitable [Laursen & Collins, 2004]. Relationship research discusses the inherent tensions between independence and closeness during adolescence for typical families [Laursen & Collins, 2004; Sillars, Koerner et al., 2005]. Before adolescence, parents are typically a primary social contact for their children and fulfill most of their social support needs. During adolescence, the child’s increased autonomy is correlated with a decrease in closeness between parent and child. For most adolescents, this close social connection is not lost, but is transferred to other relationships (e.g., peers). This last point highlights the importance of a circle of friends during the transition to increased independence. Among families with a child with cognitive disabilities, the transition to independence is typically postponed or may never fully take place. There are a number of reasons for this, which may include the child’s limitations in self-care and functional skills, the parents’ or child’s desire to remain close and connected with one another, and the absence of a supportive circle of friends to facilitate this transition.
150
Even among the two families who participated in the probe study there was a very wide difference in the daughters’ desire for independence. Margaret was constantly pushing her parents to let her stay in her apartment alone, while her parents were concerned with safety and Margaret’s potential social isolation. In contrast, Linsey’s mother Carin desired Linsey to be more self-reliant and independent, while Linsey’s main priority was to maintain a strong social connection with her caregivers. There were also different social connection goals between parents and children. Parents’ goals for their child’s social connection and inclusion fell into two categories: 1) appropriate level of social interaction with the parents, and 2) appropriate level of social involvement in the community. Regarding the first category, both parent groups in the probe study felt that their daughters communicated sufficiently with them. Linsey’s parents even sought to reduce Linsey’s level of remote communication and increase her self-reliance. The second category is focused on social involvement in the community and a social connection with a circle of friends. Parents in all phases of this research reported that maintaining a “circle of friends” was an important but difficult goal. Parents wanted to promote socially appropriate behavior to increase social acceptance. Parents hoped that technology could help their children maintain a circle of friends, and encourage socially appropriate behavior. For the young adults with cognitive disabilities, the desire for social connection centered more on sharing emotional experiences and receiving emotional support from their primary caregivers. They desired social interaction with their parents and other primary caregivers far more than they did with other people. 9.4. Design Implications Multiple design implications stem from the emotional and social aspects of this research. These design implications have relevance beyond remote communication technology, as many of the emotional and social aspects apply to various contexts in which AT is used by an individual with cognitive disabilities. The centrality of emotion in the process of providing support is an important consideration for AT in general. This research has touched on various social facets of AT use, including the
151
network of caregivers who select and evaluate the technology, the close and complex relationship between caregivers and care recipients, and the fact that much AT supports communication which is intrinsically social. 9.4.1. Considering Hot Goals in Software Requirements and Design Emotional aspects of technology use, even in user-centered and participatory design, are often ignored completely or relegated to terse assessments of “subjective experience.” This dissertation research suggests that socio-emotional motivations are a central part of familial caregiver and care recipient behavior and should be a primary consideration in the design of cognitive assistive technologies. Overlooking the emotional components in design and evaluation runs the risk of missing a critical piece of the picture. 9.4.1.1. Traditional Software Engineering Models and User Models Traditional software engineering models, user models, and UI evaluation methods focus on functional tasks, and only represent user goals directly relevant to those tasks. For example, for the design of a word processor, a functional model might use “delete a word” as a user task [John & Kieras, 1996]. The design process is mainly a translation problem of operationalizing the tasks into subtasks and ultimately into specific actions that the system should support. Systems are then evaluated based on how well they support the prescribed functional tasks. For example, task-action mappings [Payne & Green, 1987] assume that: •
All relevant user goals can be represented as tasks
•
Tasks have a structure and can be decomposed into subtasks
•
All tasks can ultimately be mapped onto actions
These models are problematic when users’ motives or goals have a significant emotional or relational component. Emotional goals cannot be operationalized into tasks or deconstructed in the same way as functional goals. Yet when emotional motives and needs are not represented, the design goals as well as the evaluation of the software are incomplete and possibly even misleading. An example from this study was Linsey’s need to know “what time it is.” When modeled purely from a 152
functional perspective, this is an information-seeking goal that is satisfied by designing the system to tell Linsey what time it is. However, Linsey’s need was not met simply by providing this information because her underlying need for emotional support and reassurance from a caregiver was not addressed. In general, when emotional and relational motives are central in users’ goals and behavior, technology that does not consider these aspects will likely not be accepted. 9.4.1.2. Affective Computing, Emotional Models and Emotional Systems In recent years there have been attempts to include emotion in conceptual models and analytic frameworks in the fields of cognitive science, affective computing, and human-computer interaction. In cognitive science the term hot cognition, coined by Robert Abelson [1963] (cited in Lodge & Taber [2005]) roots all the way back to early information processing models based on neural processes such as spreading activation. In Abelson’s model, certain concepts became emotion-laden, or hot, when the concept co-occurred with an emotion or evaluative feeling frequently or recently. More recent development in cognitive science, such as work by neuroscientist Antonio Damasio, have given a more sophisticated account of the interdependence between rationality and emotion [Damasio, 1994]. Damasio argues that rationality is embodied rather than abstract, and thus is actually a product of our emotions rather than divorced of emotion. The small research area of affective computing within computer science focuses on how emotion can be modeled in computer systems. Emotional models in affective computing represent emotion as an essential component of cognition [Picard, 1997]. Affective computing systems are often used in the areas of virtual reality (where virtual agents or avatars are designed to have human-realistic behavior), and in pedagogical and tutoring systems in which affect is used to monitor and support learning [McQuiggan & Lester, 2006]. In general, affective computing is founded on the idea that user interfaces will be better if systems can recognize, understand, and even express human-like emotion. In human computer interaction there has been a recent “turn to emotion,” illustrated by Don Norman’s most recent book Emotional Design [2004]. Norman
153
argues that humans respond to things at three levels: the visceral, the behavioral, and the reflective. The visceral level represents our “gut reaction” to something, which is where emotion plays a strong role in determining how we ultimately perceive something. McCarthy and Wright [McCarthy & Wright, 2004] argue that as technology becomes more and more pervasive, understanding the user experience must include an account of the “felt experience” of technology which includes our emotional, intellectual, and even sensual experience with it. A common element of all of the discussions of emotion described so far is that they focus on a single individual’s emotional response to objects in the world, whether the objects are teapots (as Norman discusses) or technology. They model emotion as an internal property of a single person, and do not describe the social context in which it takes place (except McCarthy and Wright, who do discuss the relational nature of technology experience). In general, emotion discussed in cognitive science and computer science is analyzed as a cognitive process inside one’s head, which can be objectively defined and measured as another type of information. Boehner et al. [2007] criticize this approach and argue for an interactional account of emotion that recognizes its situated nature and cultural context. This is particularly relevant for the research at hand, where emotional needs and goals take place between individuals in a close relationship. The way emotion is defined determines the way it is represented in design and evaluation. Table 9.2 summarizes four different approaches to incorporating emotion in affective computing and human computer interaction.
154
Type of System
Goal / Application
Method of Design and Evaluation
Empathic Systems
Sense human emotion and respond to it by adapting the user interface. Common application is in virtual reality environments and pedagogical systems.
Often data-driven, empirically gathered categories of universal, “objective” human emotion [Prendinger & Ishizuka, 2005; McQuiggan & Lester, 2006]
In virtual environments, sensing is through user’s avatar’s behavior. In physical environments, sensing is usually through physiological metrics such as heart rate and skin conductivity.
Emphasize behavior-based evaluation [Picard & Daily, 2005] (physiological change such as heart rate or skin conductivity), or change in task performance.
Emotionally Intelligent Systems
In some systems (e.g., Prendinger & Ishizuka [2005]), goal is to “fix” or “undo” human emotional response, such as stress or other emotion deemed inappropriate. Respond to events in the world with human-like emotion. Still largely in research stages, but goal is to improve the user interface in many applications, through a “virtual human” user interface [Gratch, Mao et al., 2006].
EmotionEliciting Systems
Emotionally Expressive Communication Systems
Self-reporting is also used, but is considered secondary and less “objective.”
Designed and evaluated using cognitive emotional models (which treat emotional states as cognitive states).
Invoke an emotional response (usually positive).
Designed by attending to beauty, aesthetics of objects.
Improve user experience through an aesthetic and “viscerally” pleasing design [Norman, 2004]
Evaluated by measuring task performance (assumes attractive things work better), and degree of adoption/diffusion Design methods vary.
Enhance human-human communication by conveying emotional/relational component (such as sending communicators’ facial expressions with their email messages [Angesleva, Reynolds et al., 2004])
Evaluated by effectiveness of human communication.
Table 9.2: Four approaches to incorporating emotion in technology Each of these categories treats emotion differently in their goals and design approach, which results in a different approach to evaluation. Empathic systems and 155
emotionally intelligent systems rely on emotional models that define objective, universal human emotions, and map them to cognitive and physiological states. These systems have been found to be more successful in controlled environments, where human trainers and testers can produce emotions on-demand that are clear and specific [Boehner, DePaula et al., 2007]. These systems are less effective in measuring and understanding human emotion in naturalistic environments, where emotions are subtle, mixed, and dynamic. Emotion-eliciting systems deal mainly with pleasing the user and creating a sense of satisfaction through beauty. Emotionallyexpressive communication systems focus on enhancing human-human communication. These categories will serve as a background to discuss design challenges and implications for the research at hand. In order to understand how these different approaches to incorporating emotion into technology are relevant to AT design, we will look at the emotional goals and motives of Linsey, Margaret, and their caregivers in more detail. 9.4.1.3. In Pursuit of Hot Goals: Designing for Emotionally-Motivated Tasks Unlike empathic systems and emotionally intelligent systems which attempt to model and account for the universe of human emotion, our focus is much narrower. We have observed specific types of emotional goals that frequently occur in interactions between caregivers and care recipients (who have a close relationship). These emotional goals are summarized in Table 9.3. Emotional Goal
Description
Emotional Support
Reducing negative emotions including stress, sadness, fear and anxiety. Maintaining social order, reducing negative emotions, managing stress between individuals in a relationship, and mediating emotional responses (reducing extreme highs and lows). Increasing the sense of connection between individuals in a relationship, and creating a shared experience of positive emotions.
Emotional Balance
Emotional Closeness and Co-Incidence
Table 9.3: Emotional goals observed between care recipients and caregivers A salient feature of these goals is that they are embedded in human-human relationships and only make sense within their social context. Thus, while empathic
156
systems that only evaluate physiological response may have some usefulness in detecting and communicating basic emotional changes of users, this type of evaluation ignores the primary relational aspect of the emotional goal. Similarly, Norman’s advocacy for emotional design may be useful for improving the UI, but it does not capture the interactional nature of the emotional goals. Emotionallyexpressive communication systems seem to be much closer to the mark, but there is a paucity of literature on approaches to the design and evaluation of these systems. More directly useful design and evaluation guidelines are found in social support research [Burleson, 1994; Tardy, 1994]. A number of studies, including studies with professional therapists, have found that communication that effectively conveys emotional support has the following common properties [Burleson, 1994]: 1. Focuses on support recipient. The communication is centered on understanding and describing the recipient’s feelings and perspective. 2. Evaluatively neutral. The communication does not judge the situation or support recipient, it only describes feelings and the situation. 3. Feeling-centered. The support focuses on understanding and describing the immediate causes of feelings (e.g., reactions to events), rather than focusing on properties of the events themselves. 4. Accepting and legitimizing. The communication conveys acceptance of the distressed person. It legitimates the feelings and point of view of the support recipient. 5. Promotes reflection. The communication includes a description of the emotions experienced by the recipient, to enable them to understand and reflect on their feelings from a more “distanced” perspective. 6. Genuine. The support recipient perceives the support as authentic, sincere, and genuine. These six aspects of effective emotional support could be valuable guidelines for designing AT where communication of emotional support is a component. These guidelines apply as well for tutoring systems or other environments in which the system provides support. Social support research has found that the effectiveness of 157
emotional support is dependent on the relationship between the provider and recipient, and that support from close “non-intimate” sources is less effective than support from close friends and family [Sarason, Sarason et al., 1994; Tardy, 1994]. This suggests that it may be difficult to replace human support with computer support. A more productive goal may be to facilitate and enhance human-human communication. Social support research also provides a framework to evaluate the effectiveness of emotional support communication [Burleson, 1994, p. 14], shown in Table 9.4. The evaluation includes looking not only at immediate emotional changes in the recipient, but looking as well at the long-term effects, and how the support changes the relationship between the provider and recipient. Outcome
Description
Immediate instrumental effects
How well does the emotional support reduce the recipient’s distress in the here and now?
Long-term instrumental effects
How well does the emotional support promote the development of coping strategies that enhance the recipient’s long-term ability to manage distress?
Immediate relational and identity effects
How does this change the way the support recipient thinks and feels about the provider? How does this change the way providers think and feel about themselves?
Long-term relational effects
How does the consistent use of certain types of emotional support affect the quality of the relationship between recipient and provider? How does this change the provider’s relationship with others?
Table 9.4: Evaluation framework for emotional support communication It is significant that this evaluation framework focuses on both the recipient and provider of the emotional support. Studies have shown that ineffective emotional support can damage a relationship by causing frustration for both the recipient and provider [Burleson, 1994]. This implies that providing tools that increase the quality of support between caregivers and care recipients can improve the quality of and their satisfaction in their relationships. To complicate our understanding of how to design for emotional goals, the probe study also provided an example of how a task can transform from an emotional support task to an informational (or instrumental support) task. Specifically, Margaret’s probe gave her the ability to find out exactly when the bus would arrive
158
when she was waiting at a bus stop. This transformed the situation so that a late bus did not result in Margaret’s emotional distress and panicked phone call to her parents. This suggests that in addition to designing to effectively convey emotional support, AT designers can look at ways to reduce the occurrence of stress and anxiety. While assistive technology cannot replace human caregivers in providing emotional support, it can reduce or obviate the need for it in some cases. This may be through giving an individual an increased sense of control in their environment through access to relevant information, such as the location of their bus or the location of their caregiver. Assistive technology designed to minimize stress and anxiety will also reduce demands on a caregiver to provide emotional support. As we saw in Linsey’s case, more communication is not always better: Linsey’s parents sought to reduce communication while they were at work. However, the attempt to transform an emotionally-motivated task did not always work: two examples are Linsey’s non-adoption of the time feature on her phone, and Margaret’s practice of calling her parents in the morning to emotionally connect, even though her wake-up calls were replaced with an alarm on the phone. These two examples illustrate how tasks that appear to be primarily informational in nature may in fact be emotionally and socially motivated. 9.4.2. Considering Assistive Technology’s Social Context Communication technology is an exemplar of social technology: its fundamental purpose is to facilitate social interactions between people. At another level, all assistive technology is social technology, in the sense that its perceived purpose, and its appropriation into an environment, is a function of the goals and understanding of multiple individuals (including the person using the technology and her surrounding network of caregivers). Specifically, we find three ways that assistive technology is socially situated: in the network of caregivers that are involved in selecting, matching, and evaluating the technology; in the dual user interface of the configuration and maintenance of AT; and in the relationship between the care recipient and caregivers during technology use. Designing for a social context is more complex than designing for users in isolation. Jonathan Grudin argues for the importance of developing an
159
extensive understanding of users’ social environments when designing groupware, which has similar complexities ([1993, p. 97]): The computer is happiest in a world of explicit, concrete information. Central to group activity, however, are social, motivational, political and economic factors that are rarely explicit or stable. Often unconsciously, our actions are guided by social conventions and by our awareness of the personalities and priorities of people around us, knowledge not available to the computer. Research has shown that the adoption process of AT includes a network of caregivers across the phases of technology identification, selection, and appropriation [Fuhrer, Jutai et al., 2003; Dawe, 2006]. Professionals and AT specialists are often involved early-on in identifying and selecting assistive technology, and family caregivers are often responsible for configuring the technology and appropriating it into their environment. Fuhrer et al. describe how these different groups may have different goals with the technology use [2003, p. 1244]: Those [caregiver] groups are viewed as differing to some extent on the outcomes that they value, e.g., users’ enhanced physical functioning and well-being, reduced dependence on others, enlarged sense of control over their own lives, increased options for social participation and work, or reduced consumption of health and social services. Developing an understanding for the goals and desired outcomes for the various caregiver groups involved early in the design process is crucial, because this will ultimately determine whether the technology is adopted, and how the technology is used. A second social aspect of AT is the need for designing a user interface for the primary user, in addition to one for the caregiver to configure and maintain the technology. Carmien and Kintsch [2006] discuss the critical importance of dual user interface design for any AT that requires configuration. They argue that an exquisitely designed user interface for the primary user will still fail if the caregiver is unable to configure and customize the system. A simpler and more elegant end-user interface may actually result in the need for a more sophisticated and complex configuration interface. This is one of the challenges of designing an effective metadesign environment, which is discussed further in Chapter 10.
160
A third social aspect of AT is the relationship between caregiver(s) and care recipient, during design and use of the technology. In many cases primary caregivers will be parents or other family members, so this relationship will be close and complex. In this research, we observed that parental caregivers and their children have differing priorities and desires for independence, safety, and social inclusion. Technology that is situated in this context will be inevitably taking a position in these tensions: it may solely support the goals of the caregivers or those of the care recipient, or attempt to balance between the two. Ideally, designers will incorporate goal articulation, reflection, and negotiation explicitly in the design process. To close this discussion on a positive note, the presence of a social system not only adds complexity but can enhance the robustness of technology. Humans are far more skilled at error recovery and responding to unanticipated situations than technology systems. These skills can be leveraged and can complement the strengths of computer systems, when the system is viewed as part of a socio-technical environment. Brown & Duguid [2000, p. 70] remind us that: Design needs to attend not simply to the frailty of technological systems and the robustness of social systems, but the ways in which social systems often play a key part in making even frail technology robust. When designing assistive technology, the presence of a network of caregivers should not be viewed as a liability but rather a resource to leverage and connect with when the technology inevitably breaks down. 9.5. Summary As technology designers (AT and otherwise), we need to recognize that users may have important goals that are emotional in nature. In addition to representing users’ functional goals, we need to be in pursuit of their hot goals that are emotionally and socially based. This means that: •
We need to employ methods that study the emotional and relational component of users’ tasks and goals
•
We need to design and evaluate systems with consideration to users’ hot goals
161
•
Hot goals cannot be operationalized and evaluated in the same way as functional tasks, so they will need a new type of representation
•
When we introduce technology that facilitates support, we need to understand and consider how this will affect hot goals – will it enhance or deteriorate the quality of emotional support?
162
I shall reconsider human knowledge by starting from the fact that we can know more than we can tell. – Polanyi, 1966, p. 4
10
Design as Reflective Practice: Affordances of an Evolving Probe 10.1. Introduction In this chapter I take a step back from the data analysis and implications from this research, and I analyze and discuss the methods of the study itself – in particular, the affordances of a technology probe that evolves through use, in combination with ethnographic methods. I propose that these methods could be part of an effective methodology for designing AT in general. 10.2. Benefits and Drawbacks of Technology Probes The evolving technology probe method was chosen for this research environment because it could start out with very simple functionality and become more sophisticated as the user gained proficiency. This initial simplicity was expressed as an important technology requirement by participants in the Phase 1 interviews. The evolution was inspired by the meta-design seed-evolve-reseed (SER) model shown in Figure 6.1, in which end-users incrementally add to and modify a system according to emergent needs at use-time. The nature of this evolution could inform the 163
requirements for a meta-design environment: properties of the system that evolved differently between families, or changed on a regular basis, would need to be easy to customize. While the goals of simplicity and customization requirements definition were met, I found that the evolving probe, combined with ethnographic methods, had additional unanticipated affordances. These affordances included engaging the individuals with cognitive disabilities more directly in the design process, and supporting deeper reflection for caregivers as well as a new awareness of the abilities of their care recipients. Another benefit of the probe method was that it forced attention to the real usability barriers of the technology, whether they occurred at the level of the software, platform or environment, as described in Chapter 7. 10.2.1. Increasing Participation by Individuals with Cognitive Disabilities: “Let Me Show You What I Want” In Chapter 3, I argued that it is crucial to involve individuals with cognitive disabilities in the design of their own technology, rather than solely relying on caregivers as “proxy users” to represent the needs and goals of their care recipients. I provided multiple reasons for this: •
Design methods involving designer introspection or formal user models are less effective.
•
Caregivers do not have perfect knowledge of the abilities and needs of their care recipients.
•
The goals, motivations, and desires of individuals with cognitive disabilities do not always match those of their caregivers.
Some of these issues were underscored in this research, such as the tension between independence, safety, and social connectedness discussed in Chapters 8 and 9. One example was Margaret wanting more independence while her parents were focused on ensuring her safety. We also saw the reverse, in the case of Linsey who resisted some aspects of independence that took away highly-valued social connections. These examples support the importance of involving individuals with cognitive disabilities directly in the design process.
164
Individuals with cognitive disabilities are severely underrepresented in participatory design activities and usability evaluations [Bühler, 2001; Sutcliffe, Fickas et al., 2003; Lewis, 2006]. Engaging individuals with cognitive disabilities in technology design is challenging, and typical design and evaluation methods must be adapted to the abilities of the participants [Fischer & Sullivan, 2002; Lepistö & Ovaska, 2004]. Traditional user-centered design and participatory design activities often ask users to describe previous usage scenarios or imagine future ones. This can be very challenging for this population, as was evident in this research. During the interviews and early participatory design activities, the participants with cognitive disabilities were largely disengaged and unable to attend to the conversation. This is not to say that all individuals with cognitive disabilities are unable to participate in semi-structured interviews, but that other methods should be available for those who are not able to communicate their needs that way. During the probe study, the participants with cognitive disabilities moved from passive onlookers to active participants in the design process. There seemed to be two reasons for this: the technology probe supported knowing-in-action and reflection-inaction [Schön, 1983], and the probe provided affordances [Norman, 1990], or invited interactions, that elicited an emotional connection. Individuals with cognitive disabilities, like everyone, have a great deal of knowledge that is tacit [Polanyi, 1966] and embedded in their actions, which makes self-reporting difficult [Schön, 1983]. For these individuals, this is exacerbated by a limited language ability and difficulty with abstract thought. Rather than asking users to describe previous usage scenarios or imagine future ones, technology probes allow users to interact directly with technology and effectively “show you” what they want and need. Through observing action we (the caregivers and I) discovered knowledge and reflection that we did not see through other methods. The technology probe created an environment where participants could incorporate technology into their actions, and thus incorporate it into their situated knowledge [Suchman, 1987] and reflection. The following example illustrates how technology probes supported the emergence of new requirements through reflection-in-action. One day Linsey and I
165
were at lunch, and Linsey used her phone to call her mom and then put the phone in her purse. She inadvertently clicked the screen after it was in her purse and the phone began to make audio announcements. Linsey took the phone out, handed it to me, and asked, “How do I turn it off?” It was instantly clear what Linsey was asking for and why: she needed a way to prevent tapping the screen accidentally while carrying it around to prevent unintentional operations. In the following week’s meeting, the family and I co-designed a screen lock function (see Figure 7.9) so that Linsey could carry the device in her purse without accidentally pushing buttons. Linsey quickly learned how to use the screen lock and used it judiciously. It is unlikely Linsey would have asked for this feature if she had not actually been using the device and experienced the problem in front of me, which prompted her to reflect in the moment on what she thought would help. This research uncovered how technology probes can elicit an emotional connection with individuals with cognitive disabilities, enabling them to engage more deeply with the technology and the design process. This was due at least in part to the strong emotional connection created by pictures, familiar voices, and music. Both Linsey and Margaret cited the pictures of their family as their favorite part of the phone. In early design sessions, Margaret was very quiet and uninterested in the project. During initial usability testing with the functioning probe, Margaret became immediately attached to the device when she heard it play one of her favorite Patsy Cline songs as a reminder alarm. After using the new phone for a short time, Margaret surprised her parents and me when she asked for more phone contacts. While her parents wanted Margaret to have more social connections, they believed Margaret was uninterested in talking on the phone or calling people. Initially we only included her parents and one other caregiver in her contact list. When Margaret saw the pictures on the device, she asked if we could add some of her friends and cousins, with whom she rarely communicates otherwise. Margaret’s parents then built on this further by suggesting that we add a nightly reminder for Margaret to call someone in her contact list, to increase her social connections.
166
10.2.2. Increasing Reflection by Caregivers Enabling individuals with cognitive disabilities to use technology in existing patterns of action proved to be a powerful way to observe their situated knowledge, and to support reflection-in-action about their needs with the technology. It also supported increased caregiver reflection on their requirements for the technology. During the probe study caregivers showed a deeper engagement in the design process than in pre-use participatory design activities. Like the care recipients, experiencing the technology in situated use enabled richer reflection and a stronger connection to the device for caregivers. Through this grounded reflection, features were designed to be simpler than they may have otherwise been. I will give two examples of features that became less complex through the probe study. The Memo Player on Linsey’s probe was inspired by a stationary memo device that Linsey’s mom Carin used to record a daily message for Linsey. In the Phase 1 interview, Carin described how this simple, replaceable memo recorder was a much more effective aid than the expensive, complex devices she had tried with Linsey in the past. The device had one large button to play the memo, and a smaller button to record a memo (only a single memo could be stored at a time). During early participatory design sessions, the parents suggested that we add the memo functionality to the probe. The functionality seemed like a natural fit for the probe, and having it on the probe meant that Linsey could access it anywhere, not just at home. I designed the UI and functionality to be very close to the memo device, effectively “gift wrapping” [Fischer, 1998] the memo device in the probe software. While borrowing is a recommended user interface design method [Lewis & Rieman, 1993], gift-wrapping refers to supporting a task in a new media without considering the affordances of the new media. Gift-wrapping is often cast in a negative light, because it does not allow tasks to be transformed by technological innovation. Using Ehn’s [1993] definition of design as a dialectical process between tradition and transcendence, gift-wrapping can be viewed as overemphasizing tradition and underemphasizing transcendence. In the probe study, gift-wrapping played a positive role because it supported existing practices through simple technology that did not require initial training. The functionality could then evolve through design-in-use. 167
It is not completely true that the Memo Player was purely a gift-wrapped version of the memo device. In the main screen (Figure 7.7 (A)), the play button was a picture rather than a round button. In the caregiver screen (Figure 7.7 (B)), I added the ability to listen to and set the player to play previously recorded memos (identified by a timestamp). The ability to view and revert to previously recorded memos was an underdesigned feature [Fischer & Giaccardi, 2006] in that it did not allow memos to be deleted or managed; it was added purely to inspire Carin and Linsey’s roommate Abby to think about how they might want to use and manage previous memos. In the first design meeting after the usage period started, Carin had a number of ideas about ways to enhance the Memo Player. She suggested recording a memo that could be reused for each day of the week (e.g., “Monday’s schedule”), and programming the player to automatically update to that day’s memo. She also wanted to be able to rename old memos so she could find them more easily. Halfway through the probe study (see Table 7.3 for a timeline of probe updates), Carin and Abby reflected on the memo recordings further, and decided that they wanted to record a new message each day. Carin insisted that the “freshness” of the new daily message was important for Linsey, and the memo contents changed frequently enough that it was worth recording a new daily message (although this added a burden on the caregivers). Carin and Abby noted that they almost never reverted to old memos and did not need the ability to rename or organize them. We agreed that the Memo Player would simply keep the last ten memos, and delete older ones automatically. This is an example of a feature growing in complexity and then actually becoming simpler through design-in-use. A second example occurred with Margaret’s Bus Arrival Information Application. At the beginning of the study, Margaret’s father John described wanting a sophisticated, complex system that could track Margaret’s location, track the location of the buses, and show her a graphical map display of the bus locations. John’s vision was based on a different University of Colorado research project he had learned about called Mobility-for-All [Repenning & Ioannidou, 2006]. We added a much simpler version of this functionality to the probe (described in detail in Chapter 7). The Bus Arrival Information Application (Figure 7.11) utilized an existing website with real-
168
time bus location information, and assumed Margaret’s location (since Margaret always waited for the bus at the same place on her way to and from work). At the last meeting of the probe study, John reflected that the simple screen added to the probe was really all that Margaret needed – a more complex solution was not necessary. These two examples demonstrate that features do not always increase in number and complexity on an evolving technology probe. Sometimes features that were originally designed to be more complex became simplified. This is not surprising: research has shown that consumers typically buy technology that has more features than they need, and then become frustrated when the excessive features detract from ease-of-use [Rust, Thompson et al., 2006]. The technology probe allowed users to balance complexity with ease-of-use by reflecting on what they really needed as they used it. 10.2.3. Informing Researchers’ and Caregivers’ Understanding of Care Recipients’ Abilities and Motivations Another benefit of the technology probe was that it provided insights into the abilities of the care recipients that were previously unknown by either the caregivers or me. The combination of ethnographic observation and technology probe data was a powerful way for me to gain insight into users’ skills and motivations. There are a number of examples in Chapter 8 showing how data collected through the probe and direct observation were complementary to descriptions provided by caregivers. In some cases these data even contradicted caregivers’ accounts of their care recipient’s abilities and motivations. An example that illustrates this point is when I rode the bus with Margaret to work, and developed deeper insight into her ability to tell time and ride the bus. Margaret was very comfortable and confident on the bus. When the bus arrived to her transfer point, Margaret confidently said goodbye to the bus driver and sat down on the bus stop bench. She then looked at her watch and said, “The bus will be here at about 11:05 AM.” Before this moment I had no idea that Margaret had the ability to memorize the bus schedule, nor did I know she could relate the schedule to the current time in the way she did. In an earlier interview, her father had tersely stated,
169
“Margaret and time don’t mix.” (This example illustrates the point that caregivers did not always provide complete and accurate data about the abilities of care recipients.) While waiting at the bus stop, Margaret explained that she hates when the bus is late because it makes her late for work. She described her routine when she gets to work: first, she likes to get a certain tea at the café inside the grocery store, and then she puts her bag away in the back room and starts work. When the bus is late, she does not get to have her tea and her manager will get upset if she is late. When the bus was not in view by about 11:03, Margaret started to become agitated. When she saw the bus come around the corner, she quickly stood up, put on her jacket, located her bus pass, and put her backpack over her shoulder. I was still sitting on the bench turning off my laptop and she urged me to get ready, saying “Earth to Melissa!” This experience demonstrated that Margaret already had a system in place to wait for the bus that involved her own knowledge (memorizing when the bus should arrive) as well as tools (her watch and her bus pass). She checked her watch frequently when the bus was close to arriving. She watched for the bus to come around the corner, and as soon as it was in view she stood up and collected her jacket and backpack. I learned that the bus being on time was important to her because she coveted her schedule and having her before-work routine. When the bus was late she became upset and usually called home, in part for emotional reassurance from a loved one. This led me to hypothesize that providing the bus arrival information to Margaret was important not because she did not know the bus schedule, but because it gave her increased awareness and sense of control, while reducing anxiety when the bus was late. Parents also learned about their children’s abilities during the probe study. Research has shown that caregivers often overestimate or underestimate their care recipient’s abilities [Hawkey, Inkpen et al., 2005; Carmien, 2006]. In this study, neither of the parent groups felt like they had perfect knowledge of their daughter’s abilities and behavior. For example, in the Phase 2 interview Margaret’s parents stated that Margaret was unable to use the menus on her previous mobile phone. Yet Margaret mysteriously seemed to be able to return missed calls – both parents reported this, but did not know how she did it. They hypothesized that because
170
Margaret primarily sent and received calls from her mother’s mobile phone, she always called that number when she missed a call. During one of the participatory design meetings, we revisited this question. Margaret spontaneously demonstrated that she could press the “Send” button once from the main menu on her old phone and view a list of recent calls (e.g., “Mom’s cell,” “Dad’s office,” etc.). She had learned how to move up and down in the list, and how to initiate a call by pressing “Send” again with one of the names selected. Both parents were surprised; her mother did not even know there were any contacts in Margaret’s phone. When I asked if the contacts had been added recently, the parents were not sure and guessed that their son Tom had done it. When her dad asked Margaret if her brother had helped her with her phone, Margaret replied that he had not, but that was as much information as we could gather. This illustrates how it can be difficult or impossible to get all the relevant information one needs from caregivers or through conversations with the individual with cognitive disabilities, because caregivers may not fully understand what the care recipient is doing or is capable of. The probe study was unexpectedly a process of the parents learning about their child. Chapter 8 describes numerous examples where parents learned and reflected on their children’s motivations as well as abilities, such as Margaret’s underestimated desire for social connection and reassurance. In the design meetings during the probe usage period, the parents were very interested in viewing the data visualization tool I created (described in Chapter 7), since it allowed them to learn more about how their daughter used the phone and how she dealt with the problems she encountered. The data collected through the probe gave the parents a detailed view of behavior that they usually did not have access to. 10.2.4. Focusing on Real Usability Barriers The breakdowns that occurred during the technology probe usage period were real and authentic. The findings from the probe study did not arbitrarily focus on some aspect of the system, such as the research software, which controlled usability evaluations tend to do. The breakdowns occurred with the software, platform, and
171
even with the peripherals (including the chargers and batteries). As described in Chapter 7, most of the breakdowns occurred at the platform level or “between tasks,” and would not have been identified in a typical controlled usability session. The probe approach defined and broadened the design domain to include situated use – rather than letting the researcher define and control the design domain – and focused attention on the real barriers to technology adoption. Identifying breakdowns during a participatory design study allows the researcher to leverage the participants’ expertise and strategies for solving problems. Their techniques are sometimes simpler and more effective than a technologist’s solution. Both parent groups were experts in fitting technology to their daughters through simple adaptations. For example, one of Carin’s strategies for helping Linsey plug devices into charger cables was to put a drop of nail polish on the correct side of the charger cable and the device input. Carin was familiar with using inexpensive household items to modify devices to make them more accessible and less error-prone for Linsey. She contributed to a solution to prevent Linsey from accidentally tapping the power button or releasing the storage card on the probe. 10.2.5. Drawbacks of Technology Probes There are also drawbacks to the technology probe method which should be identified and considered. First, the technology probe method is very time and resource intensive. It is not a realistic method for conducting research with a large sample. This method is more appropriate for identifying phenomena that are then explored more broadly with less resource-intensive methods. It is not trivial to create a probe that is robust and stable enough for extended use in a naturalistic environment, even if it supports only very simple functionality. This has been confirmed in other probe studies [Langdale, Kay et al., 2006]. Yet it is important that the probe be robust so that it can instill confidence in users. When problems are found, it is important they are resolved quickly. Further, for an evolving probe it is important to empower families as designers by actually implementing some of the ideas they suggest.
172
Another risk with a technology probe is that it introduces technology early in the design process. This can limit the space of design solutions by “constructing a reality” for the users. To mitigate this, it is important to create multiple versions of early design artifacts and present them to the users to compare and evaluate. Research has shown that users give more critical feedback when they are shown multiple design ideas rather than just one [Tohidi, Buxton et al., 2006]. This is another example of the importance of empowering families as designers, so they develop a belief that they can effect change and are not “stuck” with the current solution. 10.3. Toward an Assistive Technology Design Methodology The benefits outlined above suggest that a technology probe study is a valuable method to incorporate into AT design. Combined with ethnographic methods, the technology probe method addresses some of the unique challenges of this design environment, including: 1) users may not be able to participate in traditional participatory design activities that involve abstract thought and detailed descriptions, 2) users’ abilities and needs are not fully understood at design-time, and this knowledge is distributed among the users and a network of caregivers, and 3) the research setting is the family home, which has privacy issues that make extended face-to-face observation difficult. 10.3.1. Role of Technology Probe in Design Methodology Technology probe studies are becoming more popular within HCI. Indeed, probes have been adapted for different purposes and incorporated into various research methodologies. There is some concern that the original essence of the probe as an explorative, inspiring method to be used alongside other qualitative research methods has been lost, and that probes are used as an “objective” alternative to qualitative, ethnographic methods [Boehner, Vertesi et al., 2007]. In this research, the technology probe method was used as part of a larger research methodology. This methodology is predicated on two principles:
173
1. Each participant with cognitive disabilities has unique needs, abilities, and goals that must be understood in the context of her caregiver network, and whose understanding requires use of qualitative, in-situ methods. 2. Technology users with cognitive disabilities should be empowered as designers of their own technology, which requires a shift from the traditional “design-time / use-time” dichotomy to a blended “design-in-use” approach. Within this methodology, it is important to position the technology probe as primarily a method for design research, and secondarily as a method for design evaluation. The primary role of the technology probe was to provide insights into the research environment, not to evaluate a specific solution to a design problem. The probe, in combination with face-to-face interviews and observations, gave me important insights into the participants’ goals and motivations. It highlighted the important role of emotional needs in participant behavior, and suggested requirements for a future socio-technical system to support remote communication within this context. Secondarily, the probe gave me more direct design implications for AT on a mobile platform, through real-world use and evaluation. The use of multiple data collection methods in the process of design and evaluation played an important role in this methodology. Combining qualitative and quantitative data gave me a richer perspective on a complex research environment, and employing multiple methods increased the opportunities for the care recipients and caregivers to participate. The value of combining data collected through multiple methods has been confirmed in other AT design projects with individuals with cognitive disabilities [Lepistö & Ovaska, 2004; Hawkey, Inkpen et al., 2005]. It is important not to understate the importance of ethnographic methods in this research. Frequent face-to-face interviews and observations guided me in understanding how to interpret the probe data. Periods of participant observation, in which the I spent a few hours with the participants in a natural setting, provided insight into the nuances and complexities of motivations and needs, while also illuminating the complex relationships between the individuals involved. Participant observation activities were sometimes just with Margaret or Linsey, and sometimes with the families. I found that design insights from Margaret or Linsey were much 174
more “opportunistic” than typical design studies, and happened far more often during natural activities than during interviews or design meetings. For example, one of the major contributions Linsey made to the phone design was the screen lock. Linsey had been present during many design meetings and had been using the phone for four weeks, and the need to lock the touchscreen had never come up until the moment she experienced it in front of me (as described in Section 10.2.1). 10.3.2. Adapting Ethnographic Methods This research identified a few useful adaptations to traditional ethnographic methods of interviews and observations: •
Adapting semi-structured interviews: It is inappropriate to plan a 1 1/2 hour interview with a participant with cognitive disabilities to talk about an aspect of their life. It is far more effective to have an unstructured conversation with a parent and son or daughter together and let the conversation wander to topics that are engaging to the son or daughter, with the help of the parent. This seemed to be a good way to initially meet and begin to learn about the individual with cognitive disabilities.
•
Adapting participant observations: It is best to be as authentic as possible (be yourself!) when spending time with participants with cognitive disabilities. It helps to first spend time with them in the presence of caregivers, so that you can take cues from them on appropriate techniques to communicate and relate to the individual. Perhaps more than in other research contexts, resist trying to maintain a strictly “professional relationship,” because the participants themselves will not. In my research, the participants and I developed friendships, and interacted socially outside of the research context.
10.4. Understanding Requirements for a Meta-Design Environment Ultimately, the design goal of most AT projects will not be a single system, but rather a flexible solution that can easily be adapted for different users and as new needs emerge. This research, along with other studies, has underscored the need for
175
individually customized assistive technology [Sutcliffe, Fickas et al., 2003; Carmien, 2006], and for AT to support new needs over time [Fickas, Robinson et al., 2005]. Toward this goal, the methods used in this research proved to be effective in informing the creation of a meta-design environment. The evolving technology probe study identified three types of meta-design requirements: 1. Concrete examples of the ways technology would need to be modified for each user 2. Types of changes in the technology that were likely to occur on a regular basis 3. Details about how users want to make modifications The requirements identified through the probe study are summarized in Table 10.1. While this data reflects findings from only two users, it illustrates the types of requirements that can emerge in a technology probe study and suggests initial requirements to explore with a larger sample. Type of Requirement
Examples
1. Customizations for individual users
Need for optional number pad dialer on phone: • Linsey needed a simpler device that connected her with her caregivers through picture-based contacts
• Margaret was able to learn and memorize phone numbers, so she needed the additional ability to dial on a number pad
Different applications for each user based on needs and abilities: • Linsey used a memo player and a phone application • Margaret used a bus arrival information application, reminder application, and a phone application
2. Modifications that occur on a regular basis
• Ability to add and remove phone contacts • Ability to add and change images for phone contacts • Ability to add and change alarms and reminders
3. Nature and method of modifications
• Ability to easily upload pictures from a desktop computer and crop them to the correct size for a picture contact
• Ability to remotely update a memo message on the phone
Table 10.1: Types of meta-design requirements identified by the probe study Each of these three requirement categories highlights design implications for a meta-design environment. Customizations made for individual users might be supported in configuration wizards that launch the first time the device is used. Another approach is to create standard user profiles that the user can choose from, 176
based on an analysis of typical configuration settings for different types of users [Sutcliffe, Fickas et al., 2003]. Modifications that are made frequently should be more prominent and streamlined, whereas less frequent modifications can be less prominent, to keep the UI simple. Understanding the nature of modifications, at the time when the system is initially designed, is critical for the system to be effective. 10.4.1. Avoiding the Turing Tar Pit One risk of a meta-design environment is that it can be too open and configurable, such that it does not have immediate usefulness. This type of system is known as a “Turing Tar pit:” an environment where anything is possible, but nothing of interest is easy [Perlis, 1982]. This is particularly dangerous for AT design: the Phase 1 interview study and other AT design studies (such as Carmien [2006]) have found that AT requiring a great deal of configuration up front without providing immediate value is likely to be abandoned (or purchased but never truly adopted). To avoid designing a system in which anything is possible but nothing is easy, meta-designers must determine the best possible “default” settings for the out-of-thebox experience and must understand the most common ways the system will be modified. This research has shown that meta-designers can employ technology probes to conduct mediated meta-design to gather these meta-design requirements. 10.4.2. Mediated Meta-Design Mediated meta-design can be represented in a modified version of the design-time use-time diagram shown in Figure 3.1. The modified version of the diagram (in Figure 10.1) has an additional phase at the beginning of use time, during which users are empowered as designers and can change the system with the assistance of a technology expert. During this fixed study period, the technology expert helps the user adapt and evolve the technology, while gathering data on the types of changes required and how they should be accomplished. The user is supported as a designer in the iterative process of evolving the system and reflecting on the changes. The length of the fixed study period could either be a predetermined length of time, or could be
177
until the researchers achieve data saturation (when new types of changes are no longer observed).
Figure 10.1: Design-time and use-time diagram modified to include mediated meta-design phase This process can give the technology expert crucial insights into how the technology must be designed to support the dimensions of change users are likely to make. Without this data, technology developers could work on systems for decades to ensure they are able to support any type of customization, without ever making the technology available to users. Mediated meta-design as a design activity is a way to focus the efforts of technology development on the types of modifications that users will actually make. While the ultimate goal of a meta-design environment is to empower end-users to change and evolve the system without the aid of technology experts, it is likely that some changes will always require updates by system developers. For example, on a mobile phone platform a new communication medium may emerge (such as text messaging). It is unlikely that an end-user could ever modify a handset to support this type of technology change (unless the user happens to be a technology expert). Thus,
178
mediated meta-design could also be used as a periodic activity to engage end-users as reflective designers to guide major technology changes. 10.5. Summary The key message of this chapter is that to truly empower users as reflective designers, we must give them the ability to instantiate their designs and experience them in realistic ways. Otherwise, users have no way of genuinely reflecting on their design ideas. One of the highlights of the probe study was the observation that more authentic interaction with a design object (in this case the technology probe) leads to deeper engagement in the design process, and richer reflection on the design requirements, for both individuals with cognitive disabilities and their caregivers. The technology probe method has proven its potential to play an important role in an AT design methodology that comprises multiple design and research methods. Technology probes have been shown to inform the requirements of a meta-design environment by supporting research through mediated meta-design.
179
One of the symptoms of an approaching nervous breakdown is the belief that one's work is terribly important. – Bertrand Russell
11
Summary and Future Work
11.1. Summary of Approach and Results This dissertation presented a research inquiry in assistive technology for individuals with cognitive disabilities. Rather than beginning with a preconceived hypothesis about specific technology that could benefit this population, this research was fundamentally directed by the participants’ expressed needs and goals. The Phase 1 interview study identified remote communication as an important and unexplored domain of AT for individuals with cognitive disabilities. Parents described the potential of accessible mobile phones to support their goals for their children’s increased independence, social connectedness, and safety. During the Phase 2 interview study, I explored the methods, in addition to mobile phones, that families use to communicate remotely with one another, including: written notes, answering machine messages, memos, and kitchen timers. I learned how these activities were embedded in families’ practices. This understanding of families’ existing tools and patterns of communication guided the design of the technology probes used in Phase 3. By focusing on families’ tasks and goals, the design process was able to transcend 180
the functionality of a standard mobile phone in order to explicitly support the activities of a unique socio-technical environment. During the probe usage period, the technology probe evolved to more deeply reflect and illuminate families’ relationships and practices. In some instances, the technology transcended current practices and changed the way families did things (such as Margaret’s Bus Arrival Information Application). This transcendence happened organically, through a close collaboration that created a knowledge exchange between a technology expert (me) and domain experts (the families). The goal of this research was to further our understanding of how technology can play a role in increasing independence, safety, and social connectedness for individuals with cognitive disabilities. One of the most important findings towards this goal was recognizing the primacy of emotional and relational goals in the family care model. Family members’ behavior is largely motivated by emotional and relational goals, such as seeking or providing emotional support, creating emotional co-incidence, and maintaining emotional balance (as described in Chapter 9). This is relevant beyond AT design; the field of HCI needs to create better models to account for the interactional nature of peoples’ emotional goals and needs that motivate their behavior. In addition to identifying AT design implications, this research contributes to our understanding of effective participatory design methods for individuals with cognitive disabilities. Many individuals in this population are unable to contribute to design through traditional participatory design methods. This research indicated that technology probes can enable participants with low language ability to “show you” what they want through using the technology. Enabling these participants to hold and interact with an actual object, rather than asking them to describe an abstract idea, facilitated reflection-in-action and illuminated situated knowledge-in-action that was unknown even to the individual’s caregivers. Chapter 10 presents recommendations for an empirically-grounded AT design methodology that incorporates technology probes, participatory design, and ethnography to address the unique aspects of this design environment. The combination of these methods creates affordances that are relevant beyond assistive technology. For example, these methods can be used to
181
create a mediated meta-design study to identify requirements for a meta-design environment in any domain. The three phases of this research provided a number of specific AT research and design implications. These are summarized in Table 11.1. Category of Finding/Implication
Specific Examples
Discussed in Chapter(s)
Assistive Technology Interface Implications Including graphical user interface, platform and peripherals.
• Importance of creating simple AT with a
Chapter 4
low threshold and high ceiling
• Importance of portability
Chapter 4
• Usability challenges with current mobile
Chapter 5
phones (e.g., complex menu systems)
• Need to include peripherals (e.g., charger)
Chapter 5, 7
in usability evaluation
• Need to focus on platform-level and
Chapter 7
between-task breakdowns (such as accidentally tapping power button)
• Usability issues with touchscreen buttons
Chapter 7
• Importance of screen lock for touchscreen
Chapter 7
• Effectiveness of pictures, familiar voices,
Chapter 10
and music to elicit emotional connection
Socio-technical Environment Findings Including goals and motivations of users, and social or environmental considerations that affect AT usage.
• Caregivers’ desire for recipients’ increased
Design methodology implications Including understanding participants, methods and conceptual frameworks for AT design.
• Importance of recognizing AT adoption as
independence, social inclusion, and safety
• Role of remote communication in the
Chapter 4, 5, 6, 7, and 8 Chapter 5, 8
goals listed in first bullet
• Central role of emotional and relational
Chapter 9
goals in family care model
• Interconnections, tensions between
Chapter 9
independence, social inclusion, and safety
Chapter 4
a process that involves multiple caregivers at different stages
• Importance of identifying underlying
Chapter 9
emotional and social goals as well as overt functional goals
• Technology probe affordances (including
Chapter 10
engaging individuals with cognitive disabilities as active participants, supporting deeper reflection, and focusing on real usability barriers)
• Benefit of combining multiple methods and
Chapter 10
data, including participatory design, technology probes, and ethnography
• Viewing AT design as creating a meta-
Chapter 3, 6, 10
design environment
Table 11.1: Summary of research findings and design implications
182
11.2. Future Work This research was as an exploratory inquiry into a research domain that has received little attention in the HCI community. The findings from this research contribute to much-needed theoretical models and empirical methods identified by Lewis [2005] for designing and evaluating user interfaces for individuals with cognitive disabilities. There is still much work to be done. 11.2.1. Further Exploration of Design Findings Given that the technology probe study was performed with two families, it is important to further explore and expand on the design implications found in this study with a broader sample size. As well as additional probe studies with individual families, surveys and interviews should be used to understand the prevalence, among a larger population, of the issues identified through these case studies. In particular, future studies should focus on the emotional and social aspects of AT in the familybased care model, and the role of AT in lessening caregiver burden by preventing or minimizing stressful situations that create the need for emotional support. 11.2.2. Incorporation into Existing Support Systems One question that commonly comes up in this research domain is, “How can assistive technology be developed in a scalable (i.e. economically sustainable) way, given that individual users have widely ranging abilities and needs?” In the research realm, this question becomes, “How can each case study avoid creating a ‘one-off’ solution that will not contribute to a larger understanding of the needs of the population?” These are valid concerns, given that creating a well-fit AT solution is paramount. Scherer and Bodine [2006] along with others [Fickas, Robinson et al., 2005] posit that the biggest reason for both the shortage of cognitive assistive technology, and the high abandonment rate for existing assistive technology, centers around the widely ranging needs and abilities of this population. Scherer and Bodine [2006] argue that most generic solutions do not work well for this population and most specialized solutions are too hard to configure and use. She states that:
183
Fewer than 25% of people with disabilities who could be helped by assistive technology are using it … All but the simplest technologies require a thorough person-centered evaluation early in the process of selecting interventions and devices for a particular individual. [2006, p. 259] Like Scherer, Fickas et al. [2005] emphasize the importance of a custom process for matching assistive technologies with individuals. They explain that the main reason AT is abandoned among individuals with traumatic brain injury (TBI) is an initial mismatch between the technology and users’ individual abilities and goals, compounded by a lack of technology adaptation as those abilities and goals change over time. Given the importance of well-matched assistive technology, perhaps the right approach is not to pursue a “silver bullet” technology design that will work for all users, or even a meta-design environment that is only modified by untrained parental caregivers. The Phase 1 interview study found that families generally rely on simple tools with little functionality, because parents are extremely busy and simple devices do not impose the configuration and maintenance burden of complex technologies. Perhaps the right approach is to follow the example of Fickas et al. [2005], who have developed an accessible email client for TBI patients with a wide range of abilities and needs. Over 2 years, the research group has worked with 9 individuals in a long-term design and usage study that creates an individualized interface for each user, and modifies the technology periodically according to users’ changing needs. While this approach will not scale to huge numbers, the goal of Fickas’ work is not to eliminate the need for an individualized approach. Rather, the goal is to create a tool that can be used by professional rehabilitation clinicians as part of their highly individualized process of setting and meeting goals for each patient. Fickas’ group has “honed” a process over several years to develop a “routine methodology” of bringing in new participants. During the intake process the researchers and users plan goals, measure abilities, and then design the email system based on this specific and detailed information. The technology collects data during use and provides feedback on usage patterns, in addition to reporting evolving abilities and perceived needs. The design and evaluation of the email client fits into each participant’s larger goals that are then evaluated and revised periodically with a clinician. In this way, the 184
technology is integrated productively into the existing support that these individuals receive. Individuals with cognitive disabilities receive similar support through government services. In every state, each individual with disabilities who receives services is required to have an individualized care plan that describes that individual’s abilities, needs, and preferences. This plan is updated regularly, usually once a year. In the state of Colorado, this is called an Individualized Plan (IP) [State of Colorado 250510, 2005]. Each individual has a case manager responsible for writing the IP, usually in collaboration with the individual and a network of caregivers. The IP is reviewed and updated annually to reflect the individual’s changing needs and goals. The IP is updated in a meeting with the case manager, the individual with cognitive disabilities and a group of caregivers. If the individual is using AT, an AT specialist may participate in the annual re-assessment as well. This existing individualized support structure could be better recognized and leveraged by AT designers. For example, AT could be designed to support the periodic assessment of a user’s abilities and usage patterns. Just as the participant families in the probe study were very interested in the visualization tool showing how their child used the technology, AT could include a caregiver view showing usage patterns and other areas of interest. This information could contribute to the annual re-assessment process by providing additional data about a user’s abilities and behavior that may not otherwise be known. During this re-assessment, the technology could be updated to reflect changed abilities or needs as a collaborative effort between caregivers, care recipient, and AT specialists. The meta-design approach for AT could also leverage this larger cycle of developing and updating an Individualized Plan. To illustrate, I will describe a potential design and use cycle in terms of the SER model (illustrated in Figure 6.1). The assistive technology would be initially seeded through an individual evaluation involving care recipient, caregivers and an AT specialist. During usage, the technology would evolve by collecting usage data and supporting further customizations by caregivers as needed. Where appropriate, the technology could even be designed to adapt itself automatically based on changing usage (which
185
represents another area of future work). The reseeding phase would take place during the annual meeting with the community of caregivers, when the user’s needs and abilities are re-assessed. The community of caregivers could reflect on the usage data and update the technology as needed with assistance from an AT specialist. Future research should strive to understand how AT can be designed as a resource for caregivers, including parents, case managers, and AT specialists. In addition to providing direct support for the individual with cognitive disabilities, AT could integrate into caregivers’ systems of assessing abilities, setting goals, and understanding evolving needs of their care recipients. The goal in incorporating AT more closely into existing support services should be to avoid adding burden to this process, but to support and augment it. 11.3. Conclusion To close this discussion, we can circle back to the beginning and Erik Erikson’s dilemma of the “universe of one” described in Chapter 2. In the same article where he discussed the clinical patient as a “universe of one,” Erikson gave an argument for why a deeply-focused case study approach will ultimately have broader benefits: The clinician must be guided by the proposition that if he can hope to save only a small subgroup, or, indeed, only one patient, he must disregard existing statistical verdicts. For one new case, understood in new ways, will soon prove to be “typical” for a class of patients. [Erikson, 1959, p. 84] It is also my belief that addressing real problems for a few individuals is infinitely better than addressing real problems for no one. The lessons learned in careful, wellinformed case-study research can ultimately benefit many more than the participants involved.
186
References State of Colorado 2505-10. Supported Living Services (SLS) Waiver, Section 8.500.90, Colorado State Medicaid Rules. (2005) Abelson, R. P. (1963). Computer simulation of "hot" cognition. Computer simulation of personality. Tomkins, S. and Messick, D. New York, Wiley: 277-298. AbleLink Technologies. (2007). http://www.ablelinktech.com. ADA. (1990). American with Disabilities Act of 1990. American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV®-TR), American Psychiatric Press. Angesleva, J., Reynolds, C. and O'Modhrain, S. (2004). EmoteMail. ACM SIGGRAPH 2004 Posters, Los Angeles, California, ACM Press. Barbato, C. A., Graham, E. E. and Perse, E. M. (2003). Communicating in the family: An examination of family communication climate and interpersonal communication motives. Journal of Family Communication 3 (3): 123 - 148. Bennett, G., Lindgaard, G., Tsuji, B., Connelly, K. and Siek, K. (2006). Reality testing: HCI challenges in non-traditional environments. CHI '06 extended abstracts on Human factors in computing systems, Montréal, Québec, Canada, ACM Press, 1679-1682. Berg, B. L. (2004). Qualitative Research Methods for the Social Sciences. Boston, Pearson Education, Inc. Berscheid, E. and Peplau, L. A. (1983). The emerging science of relationships. Close Relationships. Kelley, H. H., Berscheid, E., Christensen, A., Harvey, J. H., Huston, T. L., Levinger, G., McClintock, E., Peplau, L. A. and Peterson, D. R., W. H. Freeman and Company: 1-19. Bird, G. and Buckley, S. (1999). Meeting the Educational Needs of Pupils with Down Syndrome in Mainstream Secondary Schools. Down Syndrome Needs and Update 1 (4): 159-174. Boehner, K., DePaula, R., Dourish, P. and Sengers, P. (2007). How emotion is made and measured. International Journal of Human-Computer Studies (IJHCS) 65 (4): 275-291. Boehner, K., Vertesi, J., Sengers, P. and Dourish, P. (2007). How HCI interprets the probes. Proceedings of the SIGCHI conference on Human Factors in computing systems, San Jose, CA, ACM Press. 187
Boyd-Graber, J., Nikolova, S., Moffatt, K., Kin, K., Lee, J., Mackey, L., Tremaine, M. and Klawe, M. (2006). Participatory design with proxies: developing a desktop-PDA system to support people with aphasia. Proceedings of the SIGCHI conference on Human Factors in computing systems, Montréal, Québec, Canada, ACM Press, 151-160. Braddock, D. (2002). Public financial support for disability at the dawn of the 21st century. Disability at the dawn of the 21st century and the state of the states. Braddock, D. Washington, DC, American Association on Mental Retardation: 65-76. Braddock, D. and Parish, S. (2002). An Institutional History of Mental Disability. Disability at the Dawn of the 21st Century and the State of the States. Braddock, D. Washington D.C, The American Association on Mental Retardation: 3-61. Braddock, D., Rizzolo, M. C., Thompson, M. and Bell, R. (2004). Emerging Technologies and Cognitive Disability. Journal of Special Education Technology 19 (4): 49-56. Brown, B. A. T., Sellen, A. J. and O'Hara, K. P. (2000). A diary study of information capture in working life. Proceedings of the SIGCHI conference on Human factors in computing systems, The Hague, The Netherlands, ACM Press, 438445. Brown, J. S. and Duguid, P. (2000). The social life of information. Boston, MA, Harvard Business School Press. Bühler, C. (2001). Empowered participation of users with disabilities in R&D projects. International Journal of Human-Computer Studies (IJHCS) 55: 645659. Burleson, B. R. (1994). Comforting messages: Significance, approaches, and effects. Communication of social support: Messages, interactions, relationships, and community. Burleson, B. R., Albrecht, T. L. and Sarason, I. G. Thousand Oaks, CA, Sage Publications, Inc.: 3-28. Burleson, B. R. and Kunkel, A. (2002). Parental and peer contributions to the emotional support skills of the child: From whom do children learn to express support? Journal of Family Communication 2 (2): 81 - 97. Button, G. and Dourish, P. (1996). Technomethodology: paradoxes and possibilities. Proceedings of the SIGCHI conference on Human factors in computing systems, Vancouver, Canada, ACM Press, 19-26. Carmien, S. (2003). MAPS: Dynamic scaffolding for independence for persons with cognitive impairments. 9th International Conference on User Modeling (UM'2003), 408 - 410. 188
Carmien, S. (2006). PhD Thesis: Socio-technical environments supporting distributed cognition for persons with cognitive disabilities. Department of Computer Science. Boulder, Colorado, University of Colorado. Ph.D. Carmien, S., Dawe, M., Fischer, G., Gorman, A., Kintsch, A. and Sullivan, J. (2004). Socio-technical environments supporting people with cognitive disabilities using public transportation. Transactions on Human-Computer Interaction (ToCHI) 12 (2): 233-262. Carmien, S. and Kintsch, A. (2006). Dual User Interface Design As Key To Adoption For Computationally Complex Assistive Technology Rehabilitation Engineering and Assistive Technology Society of North America (RESNA) Conference, Atlanta GA, USA. Carroll, J. (2004). Examining technology acceptance: When use involves more than intentions. Proceedings of the 9th Annual Congress of the UK Academy of Information Systems (UKAIS), Glasgow, Scotland. Carroll, J. (2006). Dimensions of participation in Simon's design. Design Issues 22 (2): 3-18. Cockburn, A. (2001). Agile Software Development. Reading, MA, Addison-Wesley. Cole, E. (2006). Patient-Centered Design as a Research Strategy for Cognitive Prosthetics: Lessons Learned from Working with Patients and Clinicians for 2 Decades. CHI 2006 Workshop on Cognitive Technologies, Montreal, Canada. Crabtree, A. (1998). Ethnography in participatory design. Participatory Design Conference, Seattle, Washington, Computer Professionals for Social Responsibility, 93-105. Crabtree, A. (2004). Taking technomethodology seriously: hybrid change in the ethnomethodology-design relationship. European Journal of Information Systems 13: 195-209. Crabtree, A., Hemmings, T., Rodden, T., Cheverst, K., Clarke, K., Dewsbury, G., Hughes, J. and Rouncefield, M. (2003). Designing with care: Adapting cultural probes to inform design in sensitive settings. OZCHI 2003, Brisbane, Australia, Ergonomics Society of Australia, 4-13. Csikszentmihalyi, M. (1990). Flow: The Psychology of Optimal Experience. New York, HarperCollins Publishers. Damasio, A. (1994). Descartes' error: Emotion, reason, and the human brain. New York, Putnam Publishing. Davies, R., Marcella, S., McGrenere, J. and Purves, B. (2004). The ethnographically informed participatory design of a PD application to support communication. 189
ACM SIGACESS conference on Computers and Accesibility (ASSETS), Atlanta, GA, ACM Press, 153-160. Dawe, M. (2006). Desperately seeking simplicity: How young adults with cognitive disabilities and their families adopt assistive technologies. Proceedings of the SIGCHI conference on Human Factors in computing systems, Montréal, Quebéc, Canada, ACM Press, 1143-1152. Dawe, M., Fischer, G., Gorman, A., Kintsch, A., Konomi, S., Sullivan, J., Taylor, J. and Wellems, G. (2005). Smart care: The importance and challenges of creating life histories for people with cognitive disabilities, Proceedings of the HCI International Conference (HCII), Las Vegas, July 2005: (published on CD). Duncan, H. D. (1967). The search for a social theory of communication in American sociology. Human communication theory. Dance, F. New York, Holt, Rinehart & Winston: 236 - 263. Ehn, P. (1993). Scandinavian Design: On Participation and Skill. Participatory Design: Principles and Practices. Schuler, D. and Namioka, A. Hillsdale, NJ, Lawrence Erlbaum Associates: 41-77. Engelbart, D. C. (1995). Toward Augmenting the Human Intellect and Boosting Our Collective IQ. Communications of the ACM 38 (8): 30-33. Erikson, E. (1959). The nature of clinical evidence. Evidence and inference. Lemer, D. New York. Fickas, S., Robinson, W. P. and Sohlberg, M. M. (2005). The role of deferred requirements in a longitudinal study of email. 13th IEEE International Conference on Requirements Engineering, Paris, 145-156. Fischer, G. (1998). Making Learning a Part of Life - Beyond the 'gift-wrapping' approach of technology. Lifelong learning and its impact on social and regional development. Alheit, P. and Kammler, E. Bremen, Donat Verlag: 435-462. Fischer, G. (2000). Symmetry of ignorance, social creativity, and meta-design. Knowledge-Based Systems, Special Issue on Creativity and Cognition 13 (78): 527-537. Fischer, G. and Giaccardi, E. (2006). Meta-Design: a framework for the future of end user development. End User Development. Lieberman, H., Paternò, F. and Wulf, V. Dordrecht, Springer: 427-458. Fischer, G., Giaccardi, E., Ye, Y., Sutcliffe, A. G. and Mehandjiev, N. (2004). MetaDesign: A manifesto for end-user development. Communications of the ACM 47 (9): 33-37. 190
Fischer, G. and Ostwald, J. (2002). Seeding, evolutionary growth, and reseeding: Enriching participatory design with informed participation. Participatory Design Conference (PDC'02), Malmo, Sweden, 135-143. Fischer, G. and Sullivan, J. (2002). Human-centered public transportation systems for persons with cognitive disabilities - Challenges and insights for participatory design. Participatory Design Conference (PDC'02), Malmö University, Sweden, 194-198. Fitness, J. and Duffield, J. (2004). Emotion and communication in families. Handbook of Family Communication. Vangelisti, A. L. Mahwah, N.J., Lawrence Erlbaum Associates, Inc.: 473 - 494. Fuhrer, M., Jutai, J., Scherer, M. J. and DeRuyter, F. (2003). A framework for the conceptual modelling of assistive technology device outcomes. Disability & Rehabilitation 25 (22): 1243-1251. Gaver, W. (2002). Designing for Homo Ludens. i3 Magazine (June): 2-5. Gaver, W., Dunne, A. and Pacenti, E. (1999). Design: Cultural Probes. Interactions 6 (1): 21-29. Gillette, Y. and DePompei, R. (2004). The potential of electronic organizers as a tool in the cognitive rehabilitation of young people. Neurorehabilitation 19 (3): 233-243. Golder, S. A. (2004). The keep-in-touch phone: A persuasive telephone for maintaining relationships. Proceedings of the SIGCHI conference on Human Factors in computing systems, Vienna, Austria, ACM Press, 1551. Gratch, J., Mao, W. and Marsella, S. (2006). Modeling social emotions and social attributions. Cognition and multi-agent interactions: From cognitive modeling to social simulation. Sun, R. Cambridge, Cambridge University Press: 219251. Grudin, J. (1993). Groupware and social dynamics: eight challenges for developers Communications of the ACM 37 (1): 92-105. Hawkey, K., Inkpen, K. M., Rockwood, K., McAllister, M. and Slonim, J. (2005). Requirements gathering with Alzheimer's patients and caregivers. Proceedings of the 7th International ACM SIGACCESS conference on Computers and accessibility, Baltimore, MD, ACM Press, 142-149. Holtzblatt, K., Ketola, P. and Turner, T. (2004). Techniques for designing mobile applications with customer data. Conference on Human factors and computing systems (CHI 2004), Vienna, Austria, ACM Press, 1604-1605.
191
Hulkko, S., Mattelmaki, T., Virtanen, K. and Keinonen, T. (2004). Mobile probes. Proceedings of the third Nordic conference on Human-computer interaction, Tampere, Finland, ACM Press, 43-51. Hutchinson, H., Mackay, W., Westerlund, B., Bederson, B.B., Druin, A., Plaisant, C., Beaudouin-Lafon, M., Conversy, S., Evans, H., Hansen, H., Roussel, N., and Eiderbäck, B. (2003). Technology probes: Inspiring design for and with families. Proceedings of the SIGCHI conference on Human Factors in computing systems, Ft. Lauderdale, FL, USA, ACM Press, 17-24. John, B. E. and Kieras, D. E. (1996). The GOMS family of user interface analysis techniques: comparison and contrast. ACM Transactions on Computer Human-Interaction (TOCHI) 3 (4): 320-351. Katz, J. E. and Aakhus, M., Eds. (2002). Perpetual Contact: Mobile Communication, Private Talk, Public Performance. Cambridge, Cambridge University Press. Keates, S., Clarkson, J. and Robinson, J. (2000). Investigating the applicability of user models for motion-impaired users. Proceedings of the fourth international ACM conference on Assistive technologies, Arlington, VA, USA, 129-136. Keller, H. (1957). The open door, DoubleDay. Kelley, H. H., Berscheid, E., Christensen, A., Harvey, J. H., Huston, T. L., Levinger, G., McClintock, E., Peplau, L. A. and Peterson, D. R. (1983). Analyzing close relationships. Close relationships. Kelley, H. H., Berscheid, E., Christensen, A., Harvey, J. H., Huston, T. L., Levinger, G., McClintock, E., Peplau, L. A. and Peterson, D. R., W. H. Freeman and Company. Kensing, F. and Munk-Madsen, A. (1993). PD: Structure in the Toolbox. Communications of the ACM, Special Issue on Participatory Design 36 (4): 78-85. Ketola, P. (2004). Mobile users and out-of-box experience. Human Computer Interaction Consortium (HCIC) Workshop, Winter Park, CO. Kintsch, A. and DePaula, R. (2002). A framework for the adoption and abandonment of assistive technology. BVSD SWAAAC Conference, Winter Park, Colorado, E3 1-10. Kuniavsky, M. (2003). Observing the User Experience: A Practitioner’s Guide to User Research, Morgan Kaufmann. Langdale, G., Kay, J. and Kummerfeld, B. (2006). Using an intergenerational communications system as a 'light-weight' technology probe. CHI '06 extended abstracts on Human factors in computing systems Montréal, Québec, Canada, ACM Press, 1001-1006.
192
Laursen, B. and Collins, W. A. (2004). Parent-child communication during adolescence. Handbook of family communication. Vangelisti, A. L. Mahwah, NJ, Lawrence Erlbaum Associates: 333-348. Lepistö, A. and Ovaska, S. (2004). Usability evaluation involving participants with cognitive disabilities. Proceedings of the third Nordic conference on Humancomputer interaction, Tampere, Finland, ACM Press, 305-308. Lewis, C. (2005). HCI for people with cognitive disabilities. SIGACCESS Accessibility and Computing (83): 12-17. Lewis, C. (2006). HCI and cognitive disabilities. Interactions (ACM) 13 (3): 14-15. Lewis, C. and Rieman, J. (1993). Task-centered user interface design: A practical introduction, University of Colorado, Boulder. Lindholm, C., Keinonen, T. and Kiljander, H., Eds. (2003). Mobile Usability: How Nokia Changed the Face of the Mobile Phone, McGraw-Hill. Ling, R. (2004). The mobile connection: the cell phone's impact on society. S.F., CA, Morgan Kaufmann. Ling, R. (2006). Life in the nomos: Stress, emotional maintenance and coordination via the mobile telephone in intact families. The cell phone reader: Essays in social transformation. Kavoori, A. P. and Arceneaux, N. New York, NY, Peter Lang Publishing: 61-84. Ling, R. and Yttri, B. (2002). Hyper-coordination via mobile phones in Norway. Perpetual Contact: Mobile Communication, Private Talk, Public Performance. Katz, J. E. and Aakhus, M. Cambridge, Cambridge University Press: 139-169. Lodge, M. and Taber, C. S. (2005). The automaticity of affect for political leaders, groups, and issues: An experimental test of the hot cognition hypothesis. Political Psychology 26 (3): 455-482. Mann, W. C., Hurren, D., and Tomita, M. (1993). Comparison of assistive device use and needs of home-based older persons with different impairments. The American Journal of Occupational Therapy 47 (11): 980-987. Martin, B., and McCormack, L. (1999). Issues surrounding Assistive Technology use and abandonment in an emerging technological culture. Proceedings of Association for the Advancement of Assistive Technology in Europe (AAATE) Conference. Masten, D. L. and Plowman, T. M. P. (2003). Digital ethnography: The next wave in understanding the consumer experience. Design Management Journal 14 (2): 75-81. 193
McCarthy, J. and Wright, P. (2004). Technology as experience. Cambridge, The MIT Press. McQuiggan, S. W. and Lester, J. C. (2006). Learning empathy: a data-driven framework for modeling empathetic companion agents. Proceedings of the fifth international joint conference on Autonomous agents and multiagent systems, Hakodate, Japan, ACM Press, 961-968. Moore, M. M. (2003). Real-world applications for brain-computer interface technology. IEEE Transactions on Neural Systems and Rehabilitation Engineering 11 (2): 162-165. Morris, M., Lundell, J. and Dishman, E. (2004). Catalyzing social interaction with ubiquitous computing: a needs assessment of elders coping with cognitive decline (Late Breaking Paper). Conference on Human Factors in Computing Systems (CHI 2004), Vienna, Austria, ACM Press, 1151-1154. Muller, M. J. and Kuhn, S. (1993). Participatory design. Communications of the ACM 36 (6): 24-28. Mynatt, E. D., Rowan, J., Craighill, S. and Jacobs, A. (2001). Digital family portraits: supporting peace of mind for extended family members. Proceedings of the SIGCHI conference on Human Factors in computing systems, Seattle, WA, ACM Press, 333-340. Nardi, B. A. (1997). The use of ethnographic methods in design and evaluation. Handbook of Human-Computer Interaction. Helander, M. G., Landauer, T. K. and Prabhu, P. V. Amsterdam, Elsevier Science B.V. 1: 361-366. Newcomb, E., Pashley, T. and Stasko, J. (2003). Mobile computing in the retail arena. Proceedings of the SIGCHI conference on Human Factors in computing systems, Fort Lauderdale, FL, ACM Press, 337-344. Newell, A., Carmichael, A., Gregor, P. and Norman, A. (2002). Information technology for cognitive support. The Human-Computer Interaction Handbook: Fundamentals, Evolving Technologies, and Emerging Applications. Jacko, J. and Sears, A. Mahway, New Jersey, Lawrence Erlbaum Associates: 464-481. Newell, A., Morgan, M., Gregor, P. and Carmichael, A. (2006). Theatre as an intermediary between users and CHI designers. CHI '06 extended abstracts on Human factors in computing systems, Montréal, Québec, Canada, ACM Press, 111-116. NextBus, I. (2007). http://www.nextbus.com Nieminem-Sundell, R. and Vaananen-Vainio-Mattila, K. (2003). Usability meets sociology for richer consumer studies. Mobile Usability: How Nokia Changed 194
the Face of the Mobile Phone. Lindholm, C., Keinonen, T. and Kiljander, H., McGraw-Hill Companies, Inc. Norman, D. (2004). Emotional design: Why we love (or hate) everyday things. New York, Basic Books. Norman, D. A. (1988). The psychology of everyday things. New York, Basic Books. Norman, D. A. (1990). The design of everyday things. New York, Currency Doubleday. Orlikowski, W. and Gash, D. C. (1994). Technological frames: making sense of information technology in organizations. ACM Transactions on Information Systems (TOIS) 12 (2): 174-207. Osman, Z., Maguire, M. and Tarkiainen, M. (2003). Older Users Requirements for Location Based Services and Mobile Phones. Human-Computer Interaction with Mobile Devices and Services, Springer-Verlag Heidelberg: 352-357. Palen, L. (2002). Mobile telephony in a connected life. Communications of the ACM 45 (3): 78-82. Palen, L. and Hughes, A. (2006). When Home Base Is Not a Place: Parents’ Use of Mobile Telephones. Personal and Ubiquitous Computing Journal. Palen, L. and Salzman, M. (2002). Voice-mail diary studies for naturalistic data capture under mobile conditions. Computer Supported Cooperative Work (CSCW'02), New Orleans, LA, USA, ACM Press, 87-95. Papert, S. (1980). Mindstorms: Children, Computers and Powerful Ideas. New York, Basic Books. Pascoe, C., Ryan, N., and Morse, D. (2000). Using while moving: HCI issues in fieldwork environments. ACM Transactions on Computer-Human Interaction 7 (3): 417-437. Payne, S. J. and Green, T. R. G. (1987). Task-action grammars: A model of the mental representation of task languages. ACM SIGCHI Bulletin 19 (1): 73. Perlis, A. J. (1982). Epigrams on programming. SIGPLAN Notices 17 (9): 7-13. Perry, M., O'Hara, K., Sellen, A., Brown, B., and Harper, R. (2001). Dealing with mobility: Understanding access anytime, anywhere. ACM Transactions on Computer-Human Interaction 8 (4): 323 - 347. Picard, R. W. (1997). Affective computing. Cambridge, The MIT Press.
195
Picard, R. W. and Daily, S. B. (2005). Evaluating affective interactions: Alternatives to asking what users feel. CHI Workshop on Evaluating Affective Interfaces: Innovative Approaches, Portland, OR, ACM Press. Polanyi, M. (1966). The Tacit Dimension. Garden City, NY, Doubleday. Power, M. R. and Power, D. (2004). Everyone here speaks TXT: Deaf people using SMS in Australia and the rest of the world. Journal of Deaf Studies and Deaf Education 9 (3): 333-343. Prendinger, H. and Ishizuka, M. (2005). The empathic companion: a character-based interface that addresses users' affective states. Applied Artificial Intelligence 19 (267-285). Reimer-Reiss, M. (2000). Assistive Technology Discontinuance. Technology and Persons with Disabilities Conference. Repenning, A. and Ioannidou, A. (2006). Mobility agents: guiding and tracking public transportation users. Proceedings of the working conference on Advanced visual interfaces, Venezia, Italy, ACM Press, 127 - 134. Ridgeway, C. and Johnson, C. (1990). What Is the Relationship between Socioemotional Behavior and Status in Task Groups? American Journal of Sociology 95 (5): 1189-1212. Rieman, J. (1996). A field study of exploratory learning strategies. ACM Transactions on Computer-Human Interaction 3 (3): 189-218. Riemer-Reiss, M., Wacker, R. (2000). Factors associated with assistive technology discontinuance among individuals with disabilities. Journal of Rehabilitation 66 (3). Rogers, E. L. and Escudero, V. (2004). Theoretical foundations. Relational communication: an interactional perspective to the study of process and form. Rogers, E. L. and Escudero, V. Mahwah, N.J., Lawrence Erlbaun Associates: 3 - 22. Rogers, E. M. (1995). Diffusion of innovations (4th ed.). NY, The Free Press. Rousseau, M. K. (1999). Training maintenance workers with developmental disabilities to self-manage unscheduled task demands with a pictorial pocketprompt system. Experimental Analysis of Behavior Group UK (EABG), London. Rust, R. T., Thompson, D. V. and Hamilton, R. W. (2006). Defeating feature fatigue. Harvard Business Review 84 (2): 98-107.
196
Sarason, I. G., Sarason, B. R. and Pierce, G. R. (1994). Relationship-specific social support: Toward a model for the analysis of supportive interactions. Communication of social support: Messages, interactions, relationships, and community. Burleson, B. R., Albrecht, T. L. and Sarason, I. G. Thousand Oaks, CA, Sage Publications, Inc.: 91-112. Schejter, A. and Cohen, A. (2002). Israel: Chutzpah and chatter in the Holy Land. Perpetual contact: Mobile communication, private talk, public performance. Katz, J. E. and Aakhus, M. Cambridge, Cambridge University Press. Scherer, M. J. and Bodine, C. (2006). Technology for improving cognitive function: Report on a workshop sponsored by the U.S. Interagency Committee on Disability Research. Disability & Rehabilitation: Assistive Technology 1 (4): 257-261. Schön, D. A. (1983). The reflective practitioner: How professionals think in action. New York, Basic Books. Schuler, D. and Namioka, A., Eds. (1993). Participatory Design: Principles and Practices. Hillsdale, NJ, Lawrence Erlbaum Associates. Schulze, H. (2004). MEMOS: An interactive assistive system for prespective memory deficit compensation: Architecture and functionality. ACM SIGACESS conference on Computers and Accesibility (ASSETS), Atlanta, GA, ACM Press, 79-85. Sillars, A., Koerner, A. and FitzPatrick, M. (2005). Communication and understanding in parent-adolescent relationships. Human Communication Research 31 (1): 102-128. Strauss, A. and Corbin, J. (1998). Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory, Sage Publications, Inc. Suchman, L. (1987). Plans and situated actions: The problem of human-machine communication. Cambridge, England, Cambridge University Press. Sutcliffe, A., Fickas, S., Sohlberg, M. M. and Ehlhardt, L. (2003). Investigating the usability of assistive user interfaces. Interacting with Computers 15 (4): 577602. Tardy, C. H. (1994). Counteracting task-induced stress. Communication of social support: Messages, interactions, relationships, and community. Burleson, B. R., Albrecht, T. L. and Sarason, I. G. Thousand Oaks, CA, Sage Publications, Inc.: 71-87. Todis, B., Sohlberg, M. M., Hood, D. and Fickas, S. (2005). Making electronic mail accessible: Perspectives of people with acquired cognitive impairments, caregivers and professionals. Brain Injury 19 (6): 389-401. 197
Tohidi, M., Buxton, W., Baecker, R. and Sellen, A. J. (2006). Getting the right design and the design right Proceedings of the SIGCHI conference on Human Factors in computing systems, Montréal, Québec, Canada, ACM Press, 12431252. US technology-related assistance for individuals with disabilities act (1988). Section 3.1, "The US technology-related assistance for individuals with disabilities act of 1988". Van der Hoog, W., Keller, I. and Stappers, P. J. (2004). Gustbowl: technology supporting affective communication through routine ritual interactions. CHI 2004, Vienna, Austria, ACM Press, 775-776. Verizon Wireless. (2006). http://www.verizonwireless.com. Wright, M., Marlino, M. and Sumner, T. (2002). Meta-Design of a community digital library. D-Lib Magazine 8 (5): (available online at http://www.dlib.org). Ziefle, M. and Bay, S. (2004). Mental models of a cellular phone menu. Comparing older and younger novice users. Mobile Human-Computer Interaction -MobileHCI 2004: 6th International Symposium, Glasgow, UK, September 1316, 2004. Proceedings. Brewster, S. and Dunlop, M., Springer-Verlag Heidelberg: 25-37.
198
Appendix A Appendix A:
Phase 1 Interview Questions
These interview questions focus on current and past usage of assistive technology, the role that assistive technology and regular technology plays in the lives of individuals with cognitive disabilities, and the factors that contribute to assistive technology abandonment. At the end of the interview, I will briefly describe two CLever research projects that involve handheld devices, MAPS and Mobility for All, and I will seek participants’ feedback on key features and challenges. For teachers or parents of individuals who are currently using assistive technology or have used AT in the past, the interviews will address the concepts in Rogers’ theory of diffusion to both the teachers and the parents. These concepts are: •
relative advantage – adds benefit over other types of assistance, or present circumstance
•
compatibility – how well it fits with the individual and the context in which it is adopted
•
complexity –the perceived difficulty of the technology to understand and use
•
trialability – whether it was tested before chosen; quality/quantity of trial
•
customizability (re-invention)
•
observability – to what degree are the results of an innovation visible to others?
•
support (training and maintenance)
Note that the order in which the questions appear does not necessarily reflect the order in which they will be asked. I will try to maintain a natural flow of conversation, and will incorporate questions when they appear most relevant.
199
Introduction/Grand Tour Questions:
1.
Does this student use or has used any assistive technology devices? If so, please describe the AT devices this student uses or has used that you are aware of.
2.
Considering all the student’s activities, what technology (aside from AT) does this student use?
Questions focusing on Rogers’ framework for adoption/abandonment (only applicable for students who currently use AT, or have used AT in the past. If the AT is no longer used, these questions will be phrased in the past tense). Trialability 1. How was this device chosen for the student? (Did you play a role?) 2. If yes to above, were you given the option to test this device before choosing it? If so, how long was the trial period? Relative Advantage 3. (if yes to 3, above) How did you choose this device over other devices? 4. Are there benefits for this student in using the device (vs. other devices, or unassisted)? If so, what are they? 5. Are there difficulties with using this device? If so, what arey? 6. How much impact does this device have in this student’s life? In other words, how different would your student’s life be without this technology? Consider the following areas, and rate from 1 - 10: a. at school b. at home c. in the coummunity
200
Customizability 9. In what ways has this device been modified or customized since the student began using the device? (How many hours have been spent?) 10. On a scale of 1 - 10, how hard was it to customize/modify this device? 11. Were tools and support available for this process? If so, what were they? 12. Has the customization been incremental, or was it done all-at-once? Support 13. Have you received training for the AT? 14. What kind of technical support is available? 15. What kind of support do you have for the device? For incorporating the device in classroom activities? (e.g., daily activities/child’s needs and goals)? 16. Do you feel it is adequate? Complexity 17. How well do you feel you understand the technology aspect of this device? (scale of 1 – 10, then elaborate). 1 would mean you don’t feel comfortable solving any technical problem; 10 would mean that you never have technical difficulties that you can’t address. 18. How complex is this device to use? (On a scale from 1 – 10) Compatibility 19. How appropriate is this device for this student, given your experience with this device and other devices the student has tried? (rate from 0 – 10 if possible). 20. In what ways could it be a better fit?
Focusing on technology frame: expectations and usage 21. What were your expectations about the device when you first interacted with it? 22. Have your expectations changed? If so, how? 201
23. How did you measure successful use, how do you now? (if it’s changed)? 24. Describe the main functions of this device. 25. How is the device used in the classroom? (How is it incorporated into the teaching curriculum?) 26. How has this usage changed from the time the student received the device? (Or, since you have worked with the student) – none, a little, significantly 27. How is the AT used at home? Is this complementary with how it is used in the classroom?
For students who have never used AT 28. Have you considered any AT devices for this student? 29. If so, what are they? 30. How did you learn about them? 31. What were the most important factors that you considered? 32. How did you decide not to use the device(s)? 33. Do you feel there are activities that the student does currently that could be enhanced with technology (that may or may not exist today)? 34. What do you feel are the major challenges the student would have in using assistive technology?
Introducing CLever The University of Colorado Cognitive Levers research group is in the early phases of developing a number of tools for individuals with cognitive disabilities. One of our projects is a prompting system that guides an individual through the steps of a task, e.g., grocery shopping, or riding the bus. The prompting system uses pictures and audio, and runs on a hand-held device (show a picture of the device). The steps of tasks are described in scripts that are created by caregivers of the individual (e.g., the parent or teacher). 35. Do you think this device might be appropriate for your student? 36. What do you think would be the biggest challenges with using such a device? 37. What do you think would be the most important features? 202
Another tool that we are exploring helps individuals use public transportation. It also uses a hand-held device to present the information to the user. This tool uses GPS technology and audio-visual prompts to help the user get to the bus stop, tell the user when their bus is coming and when their stop is approaching, and guide the user to his/her final destination. Another component of this system is a monitoring program for the caregiver, which allows the caregiver to visually see where the rider is on the bus system, and contact the rider if necessary. 38. Do you think this device might be appropriate for your student? 39. What do you think would be the biggest challenges with using such a device? 40. What do you think would be the most important features?
41. May I contact you again if I have further questions?
The questions to the parents are the same as above, with small changes in wording (replacing student with child, and classroom to after-school or home).
203
Appendix B Appendix B:
Phase 2 Interview Questions
I will re-introduce myself as a researcher from the University of Colorado, and explain that I am interested in talking to the families today about how they communicate with one another when they aren’t together, and understand their feelings around safety, and independence regarding their son or daughter with a cognitive disability. At the end of the interview I will describe the diary study, present the materials, and give instructions.
Section 1: Interviewing Parents Alone Theme 1: How Remote Communication is Accomplished Today
1) How do you and communicate to one another, when you aren’t together? (e.g., leave notes, recordings, phone .. ) 2) Would you like to be able to communicate more often/less often, or is this the right amount? If more or less often, can you give examples of when communication was not enough or too much? 3) Does anyone in your family have a cell phone? If so, who? 4) Does have a cell phone? If yes: a. Please describe the function of the cell phone b. Who does call on the phone? c. What is the purpose of the calls? d. Who does receive calls from? e. How often are calls made/received?
204
f. How long do the phone calls last? g. Is there concern about monthly charges getting out of control? If so, how is this managed? h. Does have difficulties using the cell phone? If so, what are they? i. Are you concerned with losing it? If so, what are your strategies for keeping track of it? If no: j. Would you like to have a cell phone? Please explain. k. If had a cell phone (depending on reasons given in previous question), how do you think it would be used? l. Do you think would have difficulties using the cell phone (if not explained in previous question)? If so, what would they be? 5) Does your family use the phone (cell phone or other phone) to communicate with one another? If yes: a. Please give a recent example of when you contacted on the phone b. Please give a recent example of when contacted you on the phone c. In general, what is the purpose of the phone calls? d. How often do you talk to each other on the phone? 6) Does use the regular phone at home? If yes: a. Please describe how the home phone is used
205
b. Who does call on the phone? c. What is the purpose of the calls? d. Who does receive calls from? e. How often are calls made/received? f. How long do the phone calls last? g. Does have difficulties using the phone? If so, what are they? h. Do you have any concerns with the phone usage? 7) Does use the phone at work? (if child doesn’t work, ask about an appropriate activity when away from home) If yes: a. Please describe how the home phone is used b. Who does call on the phone? c. What is the purpose of the calls? d. Who does receive calls from? e. How often are calls made/received? f. How long do the phone calls last? g. Does have difficulties using the phone? If so, what are they? h. Do you have any concerns with the phone usage?
Theme 2: Independence 1) When thinking of , what does the idea of independence bring to mind?
206
2) In what ways is independent? In what ways is he/she not independent? 3) In the next 5 years, do you see becoming more independent? In what ways? (example, will he/she change living environments?) 4) Do you see your role as a caregiver changing in the next 5 years? If so, how? 5) Do you see the role of ’s other caregivers changing in the next 5 years? If so, how? Will he/she have new caregivers? 6) Based on the answer to previous question: do you think you will communicate more or less? Ideally, how often would you communicate with each other? 7) What are the biggest challenges for to being independent? 8) Does play a role in independence? Please explain.
Theme 3: Safety and Monitoring 1) Is safety a concern when it comes to ? If so, please describe your safety concerns with . 2) How do you address your safety concerns? 3) Do you monitor your ’s activities while is away from home? In what ways? 4) Does play a role? Please explain. 5) Do you feel comfortable with being alone? If so, in what contexts? When don’t you feel comfortable? 6) Would you like to be able to monitor your son/daughter in ways you don’t today? Explain. 7) Are you concerned with getting lost?
207
8) How does locate himself/herself when traveling – street signs? Landmarks? Other? 9) Is capable of describing where he/she is to someone else?
Final Questions: Future Technology
1) If you were designing a tool or device that helped you communicate with your son or daughter when you two weren’t together, what would it be like? What would it enable you to do? What would it enable your son or daughter to do? 2) Would such a tool play a role in safety? How? 3) Would such a tool play a role in independence? How?
Section 2: Interviewing Child, with Parents Present
Based on information provided by the parents, the questions to the child will have to be adapted appropriately. These are the hardest questions to determine in advance. The phrasing must be appropriate for each individual, and the number and depth of questions asked will depend on the individual’s capabilities. Below are some potential questions in these three areas:
Theme 1: How Communication is Accomplished Today
1) When you are not at home, say at work , do you ever talk to Mom and Dad? 2) Do you wish you could talk to Mom and Dad more when you’re not at home? 3) Do you talk on your cell phone (or home phone) ? 4) Do you like to talk on the phone? 5) Who (else) do you talk to? 208
6) What do you talk about? 7) Do you talk to your Mom and Dad on the phone? 8) Do you wish you could talk on the phone more? 9) Have you ever lost your cell phone?
Theme 2: Independence
1) Tell me about what you do at work (or at a different community activity, away from parents) a. Who do you see there? b. Do you like it? c. What do you like best? d. Is there anything you don’t like? e. How does it feel to be away from Mom and Dad? 2) Are there things you like to do by yourself when you’re at home? When you’re not at home? What are they? 3) What do you want to do when you grow up? 4) Would you like to live in your own house someday?
Theme 3: Safety 1) What does it mean to be safe? 2) When do you feel safe? 3) Do you remember a time when you didn’t feel safe? 4) What do your parents think about you being safe? 5) Does your cell phone (if appropriate) make you feel more safe?
209
Appendix C Appendix C:
Materials from Phase 3
C.1. Voicemail Diary “Cheat Sheet” Card The voicemail diary method is described in Chapter 6. I created “cheat sheet” cards Figure C. that were printed on colorful paper and laminated. I gave each family four cards. I also gave the families double-sided sticky foam to attach the laminated cards to their wall right next to the phone. Both mothers carried a card in their purse.
Figure C.1: Phone Diary “cheat sheet” Suggesting a fixed statement when there was nothing to report (such as “no calls”) helped the participants feel comfortable calling to leave a diary message even if they felt like they had nothing to say, and reminded them to respond to each of the three questions.
210
C.2. Phone Call Journal The phone call journal activity is described in Chapter 6. I gave each family four booklets of these journal entry templates. The booklets were bound with colorful brads. The journal entry templates were different for each family, based on the remote communication purpose categories they developed. The two figures below show the Roberts family’s journal entry template (Figure C.2), and the Smith family’s journal entry template (Figure C.3).
Figure C.2: Phone call journal template for the Roberts family
211
Figure C.3: Phone call journal template for the Smith family
212
Appendix D Appendix D:
Glossary
Term
Definition
Affordance
An ability or function that a technology system facilitates or supports and makes inviting to end-users [Norman, 1990].
Assistive Technology
Defined by the ADA as “any item, piece of equipment, product, or system, whether acquired commercially or off the shelf, modified, or customized, that is used to increase, maintain, or improve functional capabilities of individuals with disabilities” [US technology-related assistance for individuals with disabilities act, 1988].
Cognitive Disability
Individuals with a cognitive disability have substantial difficulty with thought processes. This includes conceptualizing and understanding abstraction, planning and sequencing, understanding numbers and other symbols, and interpreting subtle social cues [Braddock, Rizzolo et al., 2004]. Etiologies include Down syndrome, certain genetic disorders, birth defects, traumatic brain injury (TBI), and Alzheimer’s disease. Some individuals with cerebral palsy and autism spectrum disorder, or who have experienced a stroke, also have cognitive disabilities.
Design-time
The period in traditional software design and engineering methodologies in which the technology is designed. This period is before use begins. This is contrasted with use-time, when end-users incorporate the technology into use.
Ethnography
The study of human behavior and social systems through fieldwork. The primary ethnographic methods are semi-structured or unstructured interviews and participant observation.
Family-Based Care Model
The current prevalent care model for individuals with mild to moderate cognitive disabilities in the US. In this model, the care recipient lives in the family home as a child and young adult and receives care primarily from family members. This care model replaced the institutional care model that was prominent through the 1950’s.
Independence
The degree that individuals (or systems) have autonomy to make their own decisions and agency in their actions.
213
Term
Definition
Landline
A traditional “wired” phone line in the home.
Mock-up
A rendering of a user interface. A mock-up can be a paper and pencil sketch, or a more realistic rendering in a drawing program such as Adobe Photoshop.
Participant Observation
A data collection method in ethnography and qualitative research [Berg, 2004]. In this method the researcher gathers data by participating in a naturalistic setting of interest for an extended period of time.
Participatory Design
A design methodology that includes end-users of technology as participants in design [Schuler & Namioka, 1993].
Safety
The degree that an individual is out of risk of danger. A level of safety is a basic human need that must be present in order to pursue other higher-level needs, such as independence and social connectedness.
SER Model
The seeding-evolutionary growth-reseeding (SER) development model of technology in a meta-design framework. In this model, end-users make gradual changes to software through use. This is punctuated by episodes of reseeding when designers and end-users come together and actively re-organize the system to support the evolutionary change.
Semi-Structured Interview
A data collection method in qualitative research [Berg, 2004]. A semi-structured interview has a set of interview questions that are focused on a given topic of inquiry. Unlike a completely structured interview, the semi-structured interview does not follow a specific order of questions and is more conversational in nature. This allows the interview to explore unanticipated aspects of the topic of inquiry.
Simplicity
The degree that something is easy to understand and use. In this dissertation, simplicity is treated not as a static universal property of technology, but rather as a contextual attribute that is relative to each person’s abilities. The goal for simplicity in technology design should be to create a low threshold and high ceiling [Papert, 1980].
Social Connectedness
The degree that an individual is socially and emotionally close or interdependent with others.
Socio-Technical Environment
A level of analysis that includes developing an understanding of technology function and use in the context of a social environment [Carmien, Dawe et al., 2004].
214
Term
Definition
Technology Probe
A research and design method that involves deploying a simple piece of technology that is used by end-users in a naturalistic environment [Hutchinson, 2003]. By collecting extensive usage data, a technology probe gives designers insight into design requirements by illuminating user behavior and information about the usage environment.
Underdesign
A design principle of meta-design. Underdesign does not imply less effort or forethought, but on the contrary it means that the system is carefully designed to be open and extended by end-users.
Universe of One
The research and design principle that each individual has unique needs and abilities, and must be considered as a separate case. This term was originally used to describe the uniqueness of each clinical case in psychotherapy [Erikson, 1959].
Use-time
The period in traditional software design and engineering methodologies in which the technology is used by end-users. This period is after the design phase, which is referred to as design-time.
215
Appendix E Appendix E:
Abbreviations
Term
Definition
AT
Assistive Technology
ADA
Americans with Disabilities Act
CLever
Cognitive Levers (Research Group)
CSCW
Computer-Supported Cooperative Work
CSV
Comma-Separated Values (file format and extension)
ESM
Experimental Sampling Method
GPS
Global Positioning System
HCI
Human-Computer Interaction
IP
Individualized Plan
IT
Information Technology
MAPS
Memory Aiding Prompting System
MS
Microsoft
OS
Operating System
PDA
Portable Digital Assistant
SER
Seeding – Evolutionary Growth – Reseeding
SMS
Short Message Service (also called text messaging)
TBI
Traumatic Brain Injury
UI
User Interface
XML
Extensible Markup Language
216
