Understanding and Supporting Personal Activity Management by IT ...

4 downloads 194 Views 227KB Size Report
management processes, we present the design of a software tool to support those processes and discuss findings of its usage by four. IT service workers over a ...
Understanding and Supporting Personal Activity Management by IT Service Workers Victor M. Gonzalez

Leonardo Galicia

Jesus Favela

Manchester Business School University of Manchester, UK [email protected]

Computer Science Department CICESE, Mexico [email protected]

Computer Science Department CICESE, Mexico [email protected]

ABSTRACT Many recent studies provide evidence of the challenges experienced by knowledge workers while multi-tasking among several projects and initiatives. Work is often interrupted, and this results in people leaving activities pending until they have the time, information, resources or energy to reassume them. Among the different types of knowledge workers, those working directly with Information Technology (IT) or offering IT services – software developers, support engineers, systems administrators or database managers –, experience particularly challenging scenarios of multi-tasking given the varied, crisis-driven and reactive nature of their work. Previous recommendations and technological solutions to ameliorate these challenges give limited attention to individual’s preferences and to understanding how and what tools and strategies could benefit IT service workers as individuals. Based on the analysis of characteristics of IT service work and a consolidation of findings regarding personal activity management processes, we present the design of a software tool to support those processes and discuss findings of its usage by four IT service workers over a period of eight weeks. We found that the tool is used as a central repository to orchestrate personal activity, complementing the use of e-mail clients and shared calendars as well as supporting essential aspects of IT-service work such as multi-tasking and detailed work articulation.

Keywords Multi-tasking, Personal Activity Management, Planning, Time Management, Productivity, Information Technology Workers.

1. INTRODUCTION Studies conducted within the last ten years provide vast evidence of the challenges experienced by modern knowledge workers while multi-tasking among several projects and initiatives [3-6,8,9,13]. Researchers show that in general, knowledge workers’ activities are varied, fragmented and overlapped, which forces people to limit the focus on each activity for a short period [8, 9]. These conditions are reported to leave people with a feeling of frustration, a so called, time famine, “having too much to do and not enough time to do it” [13]. Studies also report that knowledge workers try to cope with the challenges of multiPermission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage, and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. CHIMIT'08, November 14-15, 2008, San Diego, CA, U.S.A. Copyright 2008 ACM 978-1-60558-355-6/08/0011...$5.00.

tasking and hectic schedules by using different practices to manage, prioritize and focus on what they have to do [9,13]. These personal time and task management practices can be understood as a type of meta-work encompassing and focusing on the orchestration of all the tasks, projects or work the person has committed to complete [9]. Results of previous studies show that to a minor or greater extent all knowledge workers, included Information Technology (IT) service workers, engage in metawork efforts on a regular basis and the time devoted to this activity can be significant [4,6,8,9]. We understand IT service work as a particularly instance of knowledge work that characterizes those working directly with IT and offering IT services. Some examples of this kind of role are: Software developers (people designing, coding or testing software), database managers (people creating, managing, updating or backing up database servers), or support engineers (people operating or maintaining technical infrastructures). Previous research confirms that IT service workers often multi-task among several projects and tasks [3,8,9] and highlights the typical “heroism” characterizing the job, as they have to work for long hours, meet very tight deadlines and struggle with the pressure of balancing work and life commitments [3, 13]. In this paper, we analyze previous work and draw from our empirical data to consolidate a better understanding of the needs of IT service workers with regards to the personal planning and management of activities. We aim at contributing to the efforts of other researchers by discussing and analyzing the personal dimension of activity management. We present an analysis of personal planning strategies that have been revealed from ours and others’ studies and compare it with popular time-management tactics. Using both perspectives we defined a guiding framework to inform the design and development of a tool we called the Personal Activity Manager. Finally, we also present here some preliminary results on the adoption of the tool by four IT services workers, and discuss their experiences with regards the support it provides for personal task and time management.

2. THE NATURE OF IT-SERVICE WORK IT service work is characterized by a number of aspects identified by the literature. Firstly, IT service workers often experience tension to manage what can be considered proper engineering (or technical) work versus all those other tasks resulting from collaborating with colleagues and being part of an organization. Perlow highlights this tension, while describing how the participants in her ethnographic study constantly referred to having to balance between “real engineering work” and “everything else” that they do [13]. This tension can in part be the result of engineering training which often prepares people to design, analyze and test software and hardware, with less emphasis on the organizational context framing those processes.

The tension, however, can also result from a second common characteristic of IT service work: the degree of structure of their work activities. Empirical evidence shows that IT service workers carry out a mix of structured and unstructured work activities [2]. Structured activities refer to standardized processes, procedures, and tools in order to provide and guarantee high quality IT services for the clients and meet regulations established in contracts signed with them or certifiers (e.g., ISO, ITIL). In contrast, unstructured activities involve local work practices, custom developed tools, ad hoc collaborations with colleagues, and informal procedures [2]. Through unstructured activities, IT service workers obtain and share information with co-workers, and reach agreements that serve to clarify the scope and time constraints for their own work [9,13]. Both structured and unstructured activities demand collaboration with others at some points in time, and these collaborations lead to multi-tasking, a third common characteristic of IT service work. Because IT service workers have to collaborate with many people in parallel, this often results in the need to constantly switch from one activity to another [8]. Interestingly, the actual level of multi-tasking is no less intensive for IT services workers than for other professional roles. Previous studies have compared managerial, business analysts and IT workers, finding that the number of projects, and the time spent per day on each one are similar among roles [8]. Because of its nature, however, IT-service work may demand less interactive activities and more extended periods of solo work. Perlow reports participants in her study spending close to 60% of their time working alone [13]. Similarly, Gonzalez and Mark found that the software developers they studied spent longer periods of uninterrupted work on their personal computers [9]. Later on, as part of the same study, Mark et al. found that software developers and business analysts experience fewer interruptions when compared with managers [13]. They observed that because the nature of work done by developers or analysts is generally intellective, and because they interact with fewer people, the chances of work being interrupted might be reduced. The previous results might appear contradictory if one misses a fourth common characteristic of IT service work: its crisisdriven nature. The life of IT service professionals can be described as a chain of periods of solo, preplanned work, where sustained focus and concentration are essential, and periods of collaborative, intensive and inquiry-based work, where results have to be achieved immediately. Previous studies found how IT service workers experience critical situations and emergencies that suddenly drive them away from the task at hand without any chance to complete what they were doing [3,13]. Barrett and his colleagues describe how the systems administrators they studied have to deal with so called “critical situations” resulting from failures in the systems they manage for their customers (e.g., a server crash) forcing them to stop all other activities and focus on reaching a solution immediately [3]. Similarly, Gonzalez and Mark, studying a related IT-service context, found how the potential financial implications resulting from systems errors, became strong ‘magnets’ diverting software developers from whatever they were doing to bring the systems back to normal operation [8]. Thus, IT-service workers often have to deal with being able to recover from interruptions and drive through the day with their work. This demands a constant adjustment of plans, and priorities at the personal level. Providing adequate support for IT service workers and their multi-tasking needs has resulted in a number of proposals, some

based on technology and others based on implementing specific strategies or tactics. For instance, the work by Bailey, and his colleagues, aims at providing better support for the delivery of IT services with the help of a system that facilitates the management of both structured and unstructured activities [2]; similarly, Perlow comments on how strategies of negotiated time allocation introduced during her study – quite time during the morning when people were not supposed to interrupt each other –, reduced work fragmentation and maximized sustained focus on solo work [13]. One can argue that a solution to help people cope with the challenges of multi-tasking must consider the collaborative nature of IT service work; at the same time, it is also important to recognize the limits of providing generic templates or models for working, since individuals often develop their own strategies, conceptualizations and particular ways of tracking and managing their work. For instance, Bailey and colleagues comment on an IT transition manager who developed a personalized spreadsheet to track information from a number of computer servers [2]. Although they highlight this as a case where the artifact could potentially become a barrier for collaborative work – co-workers who could find it difficult to make sense of the personal artifact –, the scenario also highlights that the strategy in fact could be foremost optimal for the individual. At the personal level the spreadsheet is a useful tracking tool and, in the first place, perhaps never designed to be used by others. Crafting and personalizing personal tools is a well documented phenomenon of task management and multi-tasking [e.g. 5, 6]. Following Kidd, we can say that each knowledge worker can develop “a different internal “configuration” based on changes wrought in their thinking and outlook by the situations they have encountered, the information they have absorbed and the particular way they have made sense of these” [11]. This ‘configuration’ includes task and time management practices and the actual definition and articulation of work [13]. Consequently, and without ignoring the need for supporting collaboration, our research aims at shedding some light on the way individuals manage their work activities and providing tools and strategies that can be adequate for their needs.

3. THE ROLE OF META-WORK The efforts that people make to get things done go well beyond those specific actions directly related with achieving the goals of each commitment. Much effort is devoted to organize, plan, manage, coordinate, and retrieve all that is required to create the material and mental conditions to effectively engage in specific actions. Preparing a to-do list, reviewing a weekly schedule, creating file structures, updating a list of contacts, or checking ones computer for viruses, are all examples of what it is referred to as meta-work [8]. We can identify two distinct types of meta-work: one related to the management of resources, and the other related to the management of work activity. The management of resources is related to the provision of operational physical or digital elements to get things done. The management of activity is about the allocation of efforts where it matters, to engage into a more strategic view and management of one’s efforts. Meta-work can be understood as a phenomenon involving both situated and non-situated aspects [14]. On the one hand, the changing demands of one’s work require a constant adjustment of plans, bringing and allocating resources on demand. On the other, meta-work is a conscious effort of pre-defining work activity, sequences, and schedules. Consequently, we can say that metawork can be situated when carried out, but defined when outlined

– achieving balance between these two aspects is what makes work possible. Previous studies and empirical evidence shed light on some of the meta-work principles and processes used by people for managing their work activity, the second type of meta-work, what we refer as personal activity management. Research studies show that in general these processes are used by knowledge workers to obtain a general or partial perspective about the current status of each one of their activities, and with this perspective, get enough information to focus intelligently on those activities that demand immediate attention [3-6,9,13]. Similarly, empirical evidence, in the form of popular personal productivity tactics, recommends a number of processes which aim at achieving awareness of the roles and purposes of one’s work, as well as capturing ‘open loops’ – tasks one has committed to do – all with the objective of maintaining a sense of direction, reducing stress and increasing satisfaction [1,7]. We decided to analyze what popular personal productivity tactics have to say to our understanding of the phenomenon as they were mentioned in some previous research studies but not fully discussed [6,13]. In spite that most of those methods are not founded on formal research the undeniable popularity of some of them deserves some attention and analysis, as we can draw additional understanding of personal activity management in practice. We focused on two of the most popular methods: Stephen Covey’s First-Things-First [7], and David Allen’s Getting Things Done [1], as they emphasize different types of introspective analysis and approaches to timemanagement. Drawing from research studies and empirical evidence, we can identify three fundamental aspects of personal activity management which are discussed next. Personal activity management involves the act of capturing and listing commitments. Research on task management strategies done by Bellotti et al., identified the use of comprehensive lists of people’s tasks – task vistas – as very useful for planning purposes [6]. These lists were crafted to fit in one-page so that all to-dos could be seen together. A similar use of lists is described by Gonzalez and Mark as fundamental for what they call an overview process [9]. In order to achieve a constant focalization of activities to be done, their informants consolidated overviews of their activities, gaining knowledge about their scope and purposes, their temporal constraints, degree of development, and the next actions to conduct in each one. Other studies focusing on IT-service workers, have also pointed to list-making as one of the common processes for personal activity management [3,13]. Externalization and list making are also identified as core practices of popular personal productivity approaches. David Allen’s GTD is based on the principle of actively externalizing “to-dos” by explicitly recording them in a trusted system (a notepad, whiteboard, piece of paper) freeing the mind from remembering them [1]. On the other hand, Stephen Covey’s FTF encourages people to lists activities and externalize them, although in this case, the commitments are also scheduled [7]. Personal activity management is based on flexible definitions and execution of work commitments. Bellotti et al. report that the lists used by their participants also allowed “different perspectives for different kinds of planning” as tasks could be represented at different levels of abstraction and with different priorities [6:p6]. Their participants organized their work around projects, schedules, or collaborations and these schemes where represented on their lists. Other studies have confirmed that people prefer

listing together items with clearly different scales because this provides a holistic view of their work activity [5,9]. Personal productivity tactics offer two contrasting approaches to define work commitments. On the one hand, Covey advocates for a more structured approach with regards the framing of activity based on people’s roles, and specify prioritization and scheduling based on importance or urgency [7]. Covey encourages the identification of roles and meaningful goals to lead a more concrete objectives that can be achieved within a shorter term and contribute to reach the long term goal. On the other hand, a central element of Allen’s methodology is the organization of work at two distinct levels (personal projects vs. tasks or actions to achieve them) and the constant effort towards defining the next-action required for each project [1]. Allen also emphasizes that tasks do not have to be prioritized in any way; they can just be slotted into schedules as judged more convenient. Personal activity management involves constantly reviewing commitments. Gonzalez and Mark found that the representation of work commitments was reviewed constantly by their participants [9]. They argue that through this practice, people can maintain a state of preparedness; they can make better judgments with respect to their priorities and can move in and out of their activities as circumstances change or opportunities arise [9]. A similar process of constant reviewing was also found by Barry et al. [4]. Their study highlights the satisfaction and feeling of control people gain through the exercise of agendizing – simultaneously reviewing short and long term goals –,which is indicative of the value of prospective (what I have to do) and retrospective (what I have done) analyses of work commitments. On this regard, Covey emphasizes the importance of establishing a routine to define weekly schedules, defining priorities, regularly updating them to respond to unexpected changes, and reviewing [7]. He suggests that people have to constantly review previous week activities to see what they have accomplished and to identify where improvements can be made so that they can include these in the following weeks. Similarly, together with an emphasis on externalization of pending activity, Allen suggests processes aimed at ensuring that the person will get reminders at the right moment, and will revise their commitments with regularity [1].

4. A GUIDING MODEL FOR A PAM TOOL Previous efforts on creating meta-work tools have focused on supporting the management of resources, with systems for organizing documents, contacts, messages or the physical space [5,10,12]. Some solutions are designed to pay attention to the communication channels that give origin to activities [5], others are centered on the management of resources to facilitate the fast retrieval of work context when users switch from one activity to another [12], and some other solutions aim at creating a more natural integration between different tools and information resources [10]. Many of those efforts are framed within an area of research called Personal Information Management (PIM). Our own approach complements those efforts but focuses on supporting the meta-work required for personal activity management, which involves the articulation work that people do at the personal level, defining the what, when, and how, of each project or task one has commitment to accomplish [13].Based on the three fundamental aspects of personal activity management outlined in the previous section, we describe here a workflow model (Figure 1) used to clarify the processes and define the requirements of a tool we designed to support them. The model can also be used to guide further research and understand how

people do meta-work for managing work activity. The central idea of the workflow model is that people would move from five distinct personal activity management processes to actual execution of work commitments in an ad-hoc fashion. We discuss each element of the model in the following lines.

Figure 1. A guiding model of PAM Capture: This process involves an effort to make an explicit representation of pending activities on artifacts (physical or digital). Most common representations of activities are explicit like activity description, marks or notes write down on paper, post-it, to-do list or other artifacts. Internal factors such as personal style, or external such as work volume, influence on how or where these representations are made [4,6,8]. Furthermore, individuals do not always make explicit representations of all their activities; preferring to make just mental representations of them [4, 11]. Instances of this are activities that are easy to remember, have become routine or just a habit [6]. However, when personal cognitive limits arise and work volume or interdependence increase, the need for explicit representation can be more pressing [4]. Based on this, the process of Capture serves to clarify and articulate what has to be done and store the outcome in an artifact that functions as a “trusted system” [1]. The purpose of tools must be on making easy for people to move those pending activities and actions out of their mind into an artifact in order to review or act on them later on, whenever it is possible. Classification: This process involves the effort to provide meaning and a more detailed characterization for each one of the explicit representations of activities generated during the Capture process. The meaning of a representation can be implicitly indicated by the individuals when they represent the activity [11], but can also be made explicit declaring sub-division of the activity, classification and scheduling [3]. Tools must be flexible as some studies point out that individuals classify in different ways their activities [4, 6], using classification schemas as scope, time, priority, etc., even using very basic classification schema grouping activities as personal or work-related, or activities that belong or not to engineering work [13]. Focalization: Because it is often not possible to complete pending items within a certain period (e.g., a day), users can be selective and focus on a sub-set of the activities for which they have committed time. This subset can be defined in advance at the beginning of the day or week and determined by personal preference or constraints (e.g., pre-scheduled meetings) [8]. Studies report on people creating priority list for ensuring a nearterm execution of priority actions [6]. Thus, the Focalization

process involves the effort to focus on particular sub-set of actions or activities. This can be seen as a prioritization process because individuals select those most important or urgent actions or activities. This can also be seen as a process based on context (where the activity is carried out). Finally, it can also be seen as a process of selection based on time thinking in terms of short, medium and long-term activities [4]. In general, focalization would help individuals review the status of sub-set of activities, reprioritize and give extra attention to those incomplete actions or activities as required [4]. Tools then must provide mechanisms for sub-selection and selective visualization of pending items. Management: Activities are dynamic and evolve over time, this means that activities are likely to require new actions, or to discard or postpone others. Individuals might want to divide activities, collapse others, and move actions from one project to another. Therefore, a tool must support a Management process involving the effort to provide the adequate organization, planning and updating of pending activities. Studies show that individuals, along the workday, update the representations of activities made on artifacts [4,6,8]. These updates and modifications must be done in order to reflect changes on the information gathered about the activity and to be able to take proper and timely action when it is executed, Managing activities involve both situated and nonsituated aspects. On the one hand, changes on the activities require a constant adjustment of plans “on the fly”. But also, require a conscious effort of pre-defining work activity, sequences, and schedules. Revision: This process can be used by individuals in order to obtain a general or partial perspective of their accomplishments, make better judgments with respect to their priorities and respond to changing circumstances [8]. When individuals are doing one activity, they must often peripherally monitor the others activities and remain aware if those activities need their attention [4,8]. For this propose, tools that represent activities usually appear in a visible spot of the working space so, they can be consulted constantly [8]. With this constant revision of artifacts individuals can maintain a state of preparedness; they can make better judgments with respect to their priorities and can move in and out of their activities as circumstances change or opportunities arise, and increase the feeling of being in control of their work [4, 8]. The guiding model assumes that all the other processes emerge from the process of capturing activities. From this initial conceptualization, people engage in articulating, scheduling, prioritizing, and organizing their activities. Furthermore, the model assumes that no strict sequence has to be followed, and that not all the processes are required in all circumstances. For instance, some meta-work can be based on using the process of capturing (e.g. writing down to-do items in a list) as well as on using the revision process (reviewing the list periodically). Other people might require a focalization process (selecting and prioritizing a sub-set of items from the to-do list). In summary, the model aims at containing and integrating those processes identified by previous studies providing a consolidated set of requirement from which we can define better ways to support personal activity management.

5. THE PERSONAL ACTIVITY MANAGER Based on the processes described before and with the purpose of gaining a better understanding of the phenomenon under investigation, we designed a software application tool

called The Personal Activity Manager (PAM). The PAM tool is organized into six modules and operates in such a way that people can use any of these modules in the order and with the level of detail that they wish. The following lines describe the functionality of each module.

5.1 Capture module The module is an entry point for the users to ‘put down’, out of their mind, and into the system, all pending activities, tasks or projects they have committed or would like to do. When this module is activated, all the current work on the desktop is left in background and darkened, placing the focus of attention over the capture window and allowing users to concentrate on annotating the ideas. With the purpose of supporting spontaneous and quick capture, the module provides a simple way to input ideas and deal with the actual definition at a later stage. At this point, the entries are classified by default as ‘general’ ideas that can require, if the user wishes so, more definition and to be transformed later on into actions to do, projects, or just as an item for further reference.

5.2 Classification module The classification module allows the user to give more meaning to their entries with as much level of detail as required. Each idea is independently classified so the users can focus their attention on one entry at a time. Depending on the level of detail used to classify entries, the system would be able to provide different capabilities in terms of reporting, searching methods, and visualizations. However, detailed classification is not a prerequisite to make effective use of any module of the system. The classification module provides seven possible action categories to define an entry: Inactive, Do it as soon as possible, Scheduled, Delegated, Someday, Reference, Done, New Project. If required later on, an entry can be reclassified from most other modules.

5.3 Focalization module This module aims at facilitating users to select a sub-set of actions in order to apply a particular priority to them (e.g., mark actions as immediate attention), and with this, users can see which actions are more relevant and focus their attention on them. To facilitate the selection of actions, users can activate filters for types of actions or search them using text descriptors. Marking actions as “immediate attention” have relevance in other modules (e.g. Overview) and it is possible to apply the focalization process in other modules like Management.

5.4 Overview module This module is activated by default when the tool is invoked, presenting to users a “big picture” of all their assignments registered in the tool (Figure 2). Using this module as a review tool, users can check items already done or to be done. They can also update existing items and add new ones if required. The information is organized as a dashboard with a number of widgets. Each widget contains a particular kind of information that users might want to see including overdue actions, delegated actions, today’s tasks, tomorrow tasks, last task completed, actions for the next 7 days and recently updated items. Users can add or remove any widget to obtain as much information as they need at glance. The two main widgets in this module are the calendar and the list of actions. Through the widget calendar, users can quickly review the events scheduled for the current month (and navigate through different months back and forth). The visualization technique implemented in the calendar uses bars of different colors to identify types of scheduled events. This aims

to provide a better assimilation of information – with just a quick glance the user can know how busy they are on one particular day, week or month. By double clicking on the calendar, they can obtain more detailed views (e.g. monthly, weekly) and modify, delete or add new items. Drag and drop functionality in the calendar, let people manage events easily. The other main widget is the list of actions, here is where all items (classified or not) are located, except those classified as scheduled actions. Using this widget, users can identify all pending items or visualize a subset by using filters (e.g. only those marked as soon as possible). Users can tick-off things as they accomplish them, with those items completed appearing strikethrough.

5.5 Management module This module provides lists and tree views of all items stored in the system including actions, projects, and ideas and the user can drag one item or a set of them (e.g. a complete project) and drop it/them into a different project or context. All items, once selected, can be modified. To facilitate focusing on a specific set of entries the user can activate filters for types of projects or actions. The module provides complete control over all the items, even those that where marked as done. Users can not just add new actions or projects, but also edit important information related to them such as assigning deadlines, marking them as active or inactive and adding information notes. The module allows users to analyze in a holistic way all their projects and define the next specific actions required to move forward each one.

5.6 Review module The review module facilitates retrospective and prospective analysis of all achieved actions based on the users’ activities recorded on the tool. The module provides four main reports: 1) one listing those actions marked as done and showing (in a relative way) how many days took accomplishing each one. 2) A report for reviewing forthcoming projects and actions users can check items to be attended and how many days are left before the deadline. 3) Another report shows a summary of all projects registered on the tool. 4) The last report works as a summary of all users’ activity over a period of time (today, yesterday, last week, this week, last month, etc.).

6. RESEARCH METHODOLOGY With the purpose of improving our understanding of personal activity management, and define appropriate strategies and supportive technologies, we are conducting an evaluation study of extended use of the PAM tool. Our research aims to understand how list-making practices can facilitate personal activity management, understand the conceptualization of work activity, and understand how introspective analyses of activities can be supported with information visualization and representation tools. An initial stage of our study involved the participation of eight information workers. Among them, four participants have roles as IT service workers and the results presented here focus on those individuals; however the methodology for our study applies to all of them. All participants were recruited through our social and professional networks. Participants are all knowledge workers experiencing the challenges of multi-tasking among multiple projects, and with a disposition towards finding better ways to manage their time. The four IT service workers have the following roles: software engineer, process manager, software developer and systems project manager. The participants are asked to use the PAM tool for a period of eight weeks. At the

Figure 2. PAM tool: Overview module displaying several widgets (Overdue, Next 7 days, Today), pending items and the calendar beginning of the study, we conduct an interview and apply a questionnaire with the purpose of identifying the activities that they have to accomplish in the short and medium-term, characteristics of work context, preferences towards multi-tasking, perceptions of time control, and strategies for personal activity management. After the initial interview, the PAM tool is installed in their computers and training is provided both with regards the functional characteristics of the system, as well as the strategies to make the best of use of it. The strategies include the initial identification and collection of pending projects or tasks; the definition of subsequent actions for each project or major tasks they have to do; daily and weekly review of activities; and the practice of capturing pending issues as soon as they emerge, leaving classification for later. Halfway through the study we conduct an interview with the participants, and then conclude the study with a final interview where we measured perceptions of usefulness and ease of use about the PAM tool through a second survey. During this final interview we also discuss their experiences and changes of strategies and how and why they used the tool, and received feedback with regards the meaning of the information registered (projects, actions, ideas).To achieve our research objectives the PAM tool creates a log file which records each and every action performed by users with the use of the tool. At the beginning of the study participants are notified about this characteristic and they are informed about our methods to guarantee the confidentiality of their information. The data collected is analyzed using quantitative and qualitative techniques. We analyze the actual usage of the PAM tool by processing the log of each participant with the help of a macro that produce reports and statistics. Furthermore, we analyze the participant’s perception of the tool identifying patterns on responses and observations made during interviews and questionnaires.

7. TOM’S CASE: AN IT SERVICE WORKER AND STUDY PARTICIPANT To illustrate the actual use of the PAM tool, as well as the activities, challenges and context experienced by the IT services

workers we studied, we present here the case of Tom, one of the participants in our study. Tom is a Software Engineer working for an international company, GIT (pseudo name), and with different IT service related responsibilities as defined by the four main roles he has in the organization. Firstly, he is a Software Developer responsible of the MFSW component, part of a new version of a major software product produced by GIT. He is in charge of developing new features in his component by request or/and following a development schedule defined by project managers beforehand. Secondly, Tom is a Support Engineer in charge of fixing and documenting all problem cases reported to him by users around the world for a version of the product that includes the MFSW component. This is his main role and his priority is to fix all cases as soon as possible and in the best possible way. Consequently, as soon as he receives any new case, he puts in standby anything else he was doing and focuses exclusively on it, to figure out what the problem is, and how to fix it. The company has several priority levels for all cases reported by customers, from minor concerns to critical situations. He could receive no cases for several days but he could as well receive two or three cases in the same week. These cases could take a few hours of his time or up to several weeks. Thirdly, Tom also helps coworkers in the design of software tests. Finally, he has an “academic” role. He is involved in teaching courses to expand his company presence in the region. Tom’s multiple roles demand managing multiple and simultaneous collaborations with coworkers all over the world and personnel from other companies. Tom uses the PAM tool as part of his set of tools for reviewing and organizing his activities. He uses also an e-mail client, a shared calendar, and a paper agenda that he takes with him when he leaves the office. He uses these items regularly at different times of the day. Tom usually starts his day checking his e-mail client, the calendar application and the PAM tool. For all the assignments that come through his e-mail client that either, do not imply an immediate action or, require considerable time, he makes an entry in the PAM tool. Sometimes, he leaves items on his inbox, but marks them as unread in order to return to them

when he checks his e-mail again. He mentioned, that when he completes an activity and before initiating a new one, he usually updates his artifacts and checks the PAM tool and e-mail to know ‘what is going’ on and prioritize according to the current circumstances. For example, he said that when he is working and receives a “case” request or is interrupted with other non–urgent tasks, he tries to finish his current work or at least make a note to remember where he left the work. Tom mentioned that closing the work properly (with an annotation or similar), provides him with a sense of partial achievement, even though the task has not been completed. Tom commented that this practice reduces the feeling of frustration caused by interruptions when at the end of the day he reviews and identifies what is pending and where he left it. We noticed that each entry that Tom records on the PAM tool is described in terms of what he calls “a basic concept”: a general description of the activity, a few words that serve to remind him of the actions that he needs to perform in order to achieve the goal (e.g. “Increase SIZREX limit from 9999 to a bigger value”. Most entries are related to the Software Developer and Support Engineer roles, but he mentioned that with the use of the PAM tool he started taking a better control of those extra nonengineering activities that his company assigns him (e.g. academic role- “Meeting with director of UAG – university”). Some actions listed and stored in PAM might seem unclear to someone else (e.g. “75UTKR CPF5192 f/qwsget t/dsp80r”) but they seem clear enough for Tom to get them done. He said he uses each entry on the PAM tool as a link between the “big concept” activity, the detailed activity and all those tools necessary to perform it. The use that Tom makes of the PAM tool seems to complement the rest of his tools for personal activity management. He uses the PAM tool as a central repository to collect activities from other artifacts (e.g. e-mail message) and process them and redistribute them accordingly. Similarly, he maintains his group calendar and manually transfers appointments that he needs to have in the PAM tool, making it a key repository to orchestrate personal activity. The previous results show the particular experiences of one of our participants and the use of the PAM tool within his particular context, the following section presents aggregated quantitative results of all our participants.

8. RESULTS OF PAM TOOL USAGE This section presents results of the usage of the PAM tool by the four IT-service workers participating in our study. We report on the eight weeks of usage. The data covers 56 days (40 weekdays). Table 1 shows the average number of days the PAM tool was running on the participant’s computer. In average people used the tool for 26 days, and during these days the PAM tool ran in their computer for about half of the working day (avg. 3:28:15, s.d 2:49:18). Clearly, user 2, made the most intensive use of it, having the application open for almost the entire day (8:01:03 s.d. 6:06:44) and some days even leaving it open after leaving for home. Users 1 and 4 used the system in a more limited way.

Table 1 also shows the average time in which the PAM tool was the main window focused on the computer’s screen which was 13 minutes per day (00:12:41 s.d. 00:03:34). This can be more indicative of the real use of the tool, because this is when the user interacted with it. Interestingly, we can observe that user 4, in spite being the one with less ‘running’ time, is the one who interacted with it the most. It is important to clarify that all our participants use computers with just one display. Table 2. Percent of time that users spend on each module.

The time spent on different modules can give an initial indication of the relevance of the PAM tool’s components for the user as well of the general processes of personal activity management that we identified in Section 4. Table 2 shows the percentage of time spent on each module by each participant. We can observe that on average the Overview and the Management modules were the ones most used during the study. Clearly, Focalization and Review modules were not used much. Although it is not possible to clearly identify patterns with this number of participants, it seems that they used the tool in different ways. For instance, we can see how user 1 made relatively little use of the Management component, but was the one making the most use of the Classification module. In contrast, user 3 made no use at all of Classification, but she is the one showing more intensive use of the Management module. We believe that this is an indication of the level of flexibility provided by the PAM tool in terms of offering a number of alternatives of usage and with this allowing people to adopt whatever combination of modules they consider more appropriate for their personal needs. The limited use of Focalization can be due to the fact that the process of selecting actions and classifying them with a higher priority – allowing focusing on a sub-set of them: immediate attention actions – can be achieved from other modules. The limited use of Review can reflect a limited perceived usefulness of this functionality. Graph 1. Usefulness of the different modules of the PAM tool.

Table 1. PAM tool Usage – Time running, and on focus

Through questionnaires, we collected data about the usefulness of the different components, asking participants to provide a mark using a scale of 1-10 (top 10). We collected these marks after 4 and 8 weeks of usage. Graph 1 shows that the Overview module was the one receiving the highest score (avg.

9.75, s.d. 0.50). All modules received a better score of usefulness when users reached the 8th week, except the Review module. However, in general the results are positive, for all modules. We would have liked to see more variation to better assess the value of each component, however, considering the results, we can confirm that indeed people not just did not use the Focalization component, but its usefulness was limited (avg, 7.0 s.d. 1.41). Interestingly, although the Review component had limited use, it places, together with Management after the 4th week, as the second most useful component according to users. In relation to the number of actions and projects registered with the PAM tool, Table 3 shows that at the end of eight weeks, users registered an average of 69.50 actions (s.d. 24.77). We can see how the number of registered action per week decreased over time, from a top average of 31.25 actions (s.d. 18.87) to 3.25 (s.d. 3.40) at the end of week eight. This result was expected and it is an indication of an initial effort to capture all pending actions and then just to keep the list updated as new things came. We can notice that after week 2, the average number of new actions per week among participants is about 5 (5.5, s.d.2.9). With regards to projects, we found that over the period of our study, users recorded and kept track of an average of 8 projects (s.d. 2.16) using the tool. This figure, clearly, should not be interpreted as the total number of projects people managed during the study, just those for which they decided to use the PAM tool. As we explained later on, the PAM tool played in most cases a central role on activity management, but used in combination with other tools. Following a similar pattern as actions, most projects were created in week one and then just a few additions were made during the rest of the study.

Beyond the number of actions registered in the PAM tool it is also important to understand how many actions people actually marked as completed, because this is, after all, the main purpose of a tool like this: to get things accomplished. We found that on average about 5 actions are marked as done every week per participant. The average number of completed actions seems to decrease week after week, moving from 33.60% (1st week) to 17.70% at the end of the study. Clearly, this scenario is a result of the fact that people accumulate a backlog of actions that for some reason cannot be completed and have to be left pending for later. Interestingly, data shows a larger backlog at the middle of the study which then tends to decrease at the end. Finally, we explored the perceived value of the PAM tool with some questions (7-grade Likert scale) regarding changes on strategies, productivity, and ways of self-organizing. We applied this survey at two different times during the study, one after the 4th week of the study and again after the 8th week. Graph 2 shows the results. Perceptions of the PAM tool helping to improve personal productivity and performance, where higher at the middle of the study than at the end, but still within the boundaries of agreement, although limited. In contrast, perceptions of improvement and appropriate support from the PAM tool for organizing and planning activities increased from week 4th to week 8th and were much better than those on productivity and performance. Graph 2. Support of strategies, organization and productivity

Table 3. Total of actions registered by week.

Table 4 shows the number of actions that users associated to a particular project. Interestingly, for many users most of their registered actions were, we might call it, independent, as they were not associated with any particular project. On average, 57% of all registered actions were independent and did not require to be grouped within a higher unit of work (in our case, projects). Interestingly the actual number of actions contained within a project is not very large with just 3.5 actions per project on average. This can indicate that ‘personal’ projects do not get extremely complex in terms of number of interrelated actions, but still demand more than one action to be completed. Table 4. Total of actions associated to projects.

In summary, the analysis of the data collected from the logs presents a picture where the PAM tool is used on and off along the day, during brief periods, and helping people manage a constant stream of activity, and providing relative increments on perceived personal productivity, organization and planning. This picture, however, has to be contextualized with the actual experiences and the value of the PAM tool for the type of work done by our participants as we present in the following section.

9. THE PAM TOOL & IT-SERVICE WORK We analyze here participant’s perspectives about the value of the PAM tool to support or impact each of the four characteristics defining the nature of IT-service work as discussed in Section 2.

9.1 Work differentiation The IT-service workers participating in our study differentiate between what they perceive as their core function – engineering – versus those other activities they do such as work coordination. We found that the PAM tool was often used to manage additional activities; whereas the core ones were supported by larger proprietary company systems, as expressed by user 2:

“With the PAM tool I can control those extra activities that the company assigns to me, and in this way I prevent them to fly away from my head. Regarding my main activities like coding and support, the PAM tool functions as a check point, a reminder, because for those purposes, I have well defined systems [of the company] that I have to use.” We also found that supporting non-core work often includes using the PAM tool to keep track of personal activities, not directly related with work, but still required to organize time. This was also mentioned by user 2 during an interview: “When there is no Crit-Sit (critical situation) around here, I have spare time to organize myself… actually something that I have began to do with positive results is also registering [on the PAM tool] those other activities with no direct relation to my work… you know, going for that receipt or going to some place… this is how I begin to organize my agenda.” Although it was clear that whenever people have to use systems to do their core activities and those systems provided some level of activity management support, the PAM tool was not used for that purpose. However, following user 2, we also see that the PAM tool still plays a role as general repository of reminders of pending work (“as a check point, a reminder”).

9.2 Work structuring IT-service work often involves carrying out structured activities based on standardized processes, resulting on people having to follow well-defined procedures and tools, as well as unstructured activities defined by more informal and ad-hoc practices and tools. For some of our participants, the PAM tool clearly supports the latter type of activities, but also makes possible to achieve a more appropriate level of activity articulation which is not possible to specify using the systems supporting structured activities. User 4 mentioned this when commenting the use of MS Project vs. the PAM tool when describing the tools used for personal organization: “My main tool is MS Project, because it contains well defined activities. I have three different granularity schemas: First there is a master calendar that comes from MS Project, where I get requirements and all information that I need for creating a particular report. I mean, first I have a calendar, the MS Project that defines all the checkpoints. After that, comes the PAM tool where I define activities to do according to the things customers want, not too detailed, but more detailed than MS Project” Similarly, for user 3, the PAM tool provides details that define at the personal level what they are required to do: “On the PAM tool I write down the basic concepts of the stuff [coding and support] that I need to put attention to. Afterward, the PAM tool is like my link to go straight to the details of what I am going to do”.

9.3 Multi-tasking work The need for multi-tasking among different activities as a result of frequent interruptions and managing multiple collaborations was clearly experienced by our participants. The value of the PAM tool for coping with constant multi-tasking is that it facilitates the resumption of pending activity and attention management, as mentioned by user 3: “My eyes go directly to the Overdue widget, because plans tend to change, there are new assignments always, but overdue activities are the priority, those are ‘the today thing’, but not all

are urgent. Sometimes I even change the date of those overdue activities because I want them to appear on today widget.” We notice that for users managing people and delegating work to them, the PAM tool also works as reminder of pending activity with those collaborators, as mentioned by user 1: “I use the PAM system because it reminds me with a single word what I have to do or with whom I need to talk to, mark my priority activities or not. It is good to have control of whom you delegate something to”

9.4 Crisis-driven work Many scenarios of IT-service work result on situations where people have to take immediate action to solve problems. We referred to this as the crisis-driven nature of IT-service work. We found that the support offered by the PAM tool for these scenarios is very limited. People would not use the tool during the event, but just use it, if required, to follow-up: “When activities are bug-related I do not make any entries on the PAM tool because when this happened usually it is urgent or is a Crit-Sit (critical situation). But if customers send me an email about changes in the layout, colors or something particular then that will generate an entry on the PAM tool.” User 4 Clearly, if the issue is not solved the same day, it is possible that people will use the PAM tool to place a reminder about it: “If there is an urgent issue, all other activities are postponed; if this issue spans about 15min, then the PAM tool is not appropriate. But if this issue spans all day, then I make a note on the PAM tool with the details, and when I am back the next morning, I can check it and see what I was working on.”

9.5 The practices for using the PAM tool Beyond these particular aspects supporting (or not) the characteristics of IT-service work, how do participants use the PAM tool? We found that in essence, people adopt the PAM tool as a central repository, but not unique container, to represent their pending activities. User 2, for instance, mentions how the PAM tool helps her orchestrating activities and the use of other tools: “In the past, I did not use a common repository for all my activities, but I used several and I constantly checked them. Now all my activities are registered on the PAM with exceptions of some scheduled activities that I leave on Lotus Notes to avoid duplicate entries. However, sometimes those entries are also stored on the PAM system and from there weekly I update my personal agenda [paper]” We also noticed that the PAM tool is a repository of pending activity which does not necessarily works telling people exactly what to do, but it allows them to check for conditions on whether the pending activity can be carried out at the moment. User 1 mentions this aspect: “I don’t register everything I have to do… create that file, send that e-mail…what I really register in the PAM is the task that includes all the activities that I have to do; the big concept that functions as reminder …every day I see the concept and think, this can not happen yet, but we need to do this and this…” Finally, we observed that the ecology build by our participants to support personal activity management includes the PAM tool, their calendar (collaborative) and their e-mail client. These three tools allow them to organize themselves and manage their time appropriately.

10. DISCUSSION Supporting personal activity management by IT-service workers requires an understanding of the varied, situated, intense and fragmented nature of their work. Our goal building the PAM tool has been to understand individual practices and provide a solution for IT service workers giving them more control over their time, becoming more proactive, within the limits imposed by their hectic schedules and their collaborative context. We understand the collaborative dimension of IT service work, but also considered important to analyze and support the personal dimension. Our findings highlight the importance of taking that stance. We found that some participants kept lists of actions with descriptions that are very difficult to understand for an external eye. Among the actions registered, we could successfully understand only 65% This confirms previous studies that found that task descriptors are brief, general and personal [6]. We argue that such level of abstraction and articulation of work [14] is required and should be part of any solution for activity management (e.g. [2]). The challenge is how to provide these multiple levels of abstraction and easily move from the personal to the collaborative level and vice versa. It is important to understand that most of our results are shaped by the initial investment the person has to make to adopt a new way of doing things. Participants put considerable effort on registering their pending actions and projects, and then updated them as required. Previous studies clearly show that meta-work requires regular investment in terms of time [4,5], but our results made clear that a considerable investment is required as a entry-cost when using list-making strategies, because this is the only way that people can consolidate a full –vista, overview- of their activity. We also noticed that in spite that the PAM tool is not fully integrated with other important meta-work tools such as e-mail clients or shared calendars, nevertheless people started to use it a central repository, copying and pasting information and creating new entries as required. Of course, this is not the optimal solution, as it could be seen in the case of Tom, many pending tasks are left in the other tools (e.g. e-mail client). This makes more difficult to consolidate overviews as people have to check different tools. A potential solution is that people can directly forward e-mail messages and create entries in the PAM tool automatically. We have plans to release and evaluate this functionality in the near future. We noticed that some of the actions and projects listed by our participants have a more personal nature (e.g. “Look after the kids while – [Wife]’s meeting”). Although we did not restrict the use of the tool to work-related activity and encouraged participants it for any aspect of their life, it seems that most registered actions are work-related. We think that this is in part because people did not have an easy way to carry the PAM’s information with them. It is possible to copy the database file and take it with them to home. However, none of our participants did it. Perhaps solutions such as a Web-based interface, which can be accessed from a mobile device or a computer at home, would change the usage.

11. CONCLUSIONS AND FUTURE WORK In this paper we presented an analysis of IT-service work practices, discussed the relevance of time and task management for this type of work, and presented a guiding model of processes related to personal activity management. Using the model, we designed a support tool, and preliminary findings of its use by four IT-service workers, were discussed, highlighting some

benefits of the tool as well as its limitations. Although our findings are preliminary and we have observed a moderated use of the tool, we tend to believe that people are using it to the extent that it is required. We found that our participants used the tool, for about 15 minutes per day. A previous study found that knowledge workers spent about 44 min a day in meta-work [8]. We can assume that because the PAM tool was central, but not the only tool used by participants, this usage is reasonable. Furthermore, it is interesting to observe that when participants interact with the tool, it is mostly with the Overview component. We need to conduct further analysis to understand how the different modules are used over the day and the weeks. Future work for this project will involve the analysis of data and refinement of PAM’s functionality to explore different ways to present information for analysis and planning.

12. REFERENCES [1] Allen, D. (2001). Getting Things Done: The Art of Stress-Free Productivity. New York: Penguin Putnam.

[2] Bailey, J., Kandogan, E., Haber, E., and Maglio, P. P. (2007). Activity-based management of IT service delivery. CHIMIT. ACM Press.

[3] Barrett, R., Kandogan, E., Maglio, P. P., Haber, E. M., Takayama, L. A., and Prabaker, M. (2004). Field studies of computer system administrators: analysis of system management tools and practices. CSCW. ACM Press. pp. 388-395.

[4] Barry, D., Durnell, C., and Carroll, S. (1997). Navigating the garbage can: How agendas help managers cope with job realities. The Academy of Management Executive. vol. 11, no. 2, pp. 26-42.

[5] Bellotti, V., Ducheneaut, N., Howard, M., and Smith, I. (2003). Taking email to task: the design and evaluation of a task management centered email tool. CHI. ACM Press. pp. 345-352.

[6] Bellotti, V., Dalal, B., Good, N., Flynn, P., Bobrow, D., and Ducheneaut, N. (2004). What a To-Do: Studies of Task Management Towards the Design of a Personal Task List Manager. CHI. pp. 735742.

[7] Covey, S. (1990). Seven Habits of Highly Effective People. Edit. Simon & Schuster.

[8] Gonzalez, V. and G. Mark. (2004). Constant, constant, multi-tasking craziness: managing multiple working spheres. CHI. ACM Press. pp. 113-120.

[9] Gonzalez, V. and G. Mark. (2005). Managing currents of work: Multi-tasking among multiple collaborations. ECSCW, Springer. pp. 143-162.

[10] Kaptelinin, V. (2003). UMEA: Translating Interaction Histories into Project Context. CHI. ACM Press. pp. 353-360.

[11] Kidd, A. (1994). The marks are on the knowledge worker. CHI. ACM Press. pp. 186-191.

[12] Morteo, R., Gonzalez, V., Favela. J. and Mark, G. (2004). Sphere Juggler: Fast Context Retrieval in Support of Working Spheres. Proceedings of ENC Mexican International Conference in Computer Science. IEEE Press. pp. 361-367.

[13] Perlow, L. A. (1999). The Time Famine: Toward a Sociology of Work Time. Administrative Science Quarterly. vol. 44, no. 57-81.

[14] Strauss, A. (1988). "Articulation of Project Work: an organizational process." Sociological Quarterly 29(2): 163-178.

[15] Suchman, L. A. (1987). Plans and Situated Actions: the Problem of Human-Machine Communication. Cambridge University Press.

Suggest Documents