Elaborating distraction-conflict theory towards an analytical model for evaluating collaboration in shared virtual environments. Maria Spante Department of Technology and Society Chalmers University 412 96 Gothenburg Email:
[email protected]
1. Introduction Collaboration in virtual environments is not only a matter of how the technology supports collaboration (Hindmarsh et.al. 1998). It also matters how the persons involved can handle the social situation of collaborating together (Heldal 2003). This highlights the importance of including aspects of human-human collaboration into the technical system in which it takes place. Others have made this point. To give some examples: Biocca and Harms (2002) use models from social cognition to provide an analytical model for examining and measuring social presence in mediated situations. Blascovich (2002) uses models from social psychology to provide insights into how the mere presence of others influences our behaviour in virtual environments as well as in face to face situations. Schroeder (2002) uses the framework off Goffman from an interactionist point of view to analyse social interaction in shared virtual environments (SVEs). Thus to implement theories developed outside the SVE field is no news to the practitioners inside the field. However, there is still a need for a model that can put explicit focus on some of the particular issues that are identified when we work together- and not just to look generally at social interaction. Distraction-conflict theory developed from face to face experimental work with assignments in small groups has identified how the split focus of attention between task and an other person can create a conflict for the individual in the group, which, in turn, has an impact on how tasks becomes solved (Baron 1986). As we shall see, this knowledge becomes useful to start with while building an analytical model for evaluating collaboration in the context of shared virtual environment.
1.1.
Aim
This article will put forward an analytical model based on distraction-conflict theory that focuses on particular issues when studying collaboration in shared virtual environments.
1.2.
Overview of the article
The structure of the paper is the following. First comes a brief description of the analytical model from distraction-conflict theory. Then follows a description of the method used for elaborating the model, as well as a presentation of a study of long-term usage of networked immersive projection technology systems (IPTs, also known as walk-in virtual reality systems or CAVE systems, Cruz-Neira et.al. 1992) that served as the empirical material for elaborating the model. Next, lessons learnt from applying the distraction-conflict model on the study will be presented. Finally, an extended distraction-conflict theory model will be described.
2. Distraction conflict theory – a brief description In experimental research when people are working together to solve a task, Baron (1986) showed that individuals sometimes experience a conflict between paying attention to the task they are supposed to solve together and the person they are interacting with. That means that when two people meet in a problem-solving task, they have to cope with this distraction to be able to reach a solution for the problem at hand. According to the theory, they can react with increased arousal, meaning that they will experience some sort of emotional reaction that can facilitate the work process, but sometimes also hinder it. We can see that Baron (1986) identified how the focus of attention is guided towards the co-actor and the task at hand in a collaborative
situation (the importance of looking at focus of attention in shared virtual environments has been identified by Schroeder 2002). The theory is nicely presented by Hogg and Vaughan (1998:241):
Diagram 1. Model of Distraction-conflict theory of social facilitation Tendency to pay attention to co-actor
Individual performing a task
Presence of coactor
Attentional conflict
Increased arousal /drive
Social facilitation effect
Tendency to pay attention to task
The model starts with an individual performing a task together with another person i.e. a co-actor. When they work together, both individuals split their attention between the task and the co-actor. This split of attention can create an attention conflict that leads to a reaction inside the individual i.e. an arousal/drive that will effect the individuals performance. The model posits a social facilitation effect, meaning that performance will be better or faster for example. However, it is not always the case that the arousal will “facilitate” behaviour. Sometimes the individuals’ performance is influenced negatively with other people present (see for example Berkowitz 1988:276-279) a process often called social inhibition (Blascovich 2002:138-139). Still, for the purpose of elaborating the model in order to look explicitly at several important issues when people work together in shared virtual environment this model will be useful to build on.
3. Method and Study description 3.1.
Method for elaborating the model
To make use of the model based on distraction-conflict theory in shared virtual environments it must be tested in relation to findings. In practice, applying the model to a situation when people collaborate in a shared virtual environment will make it possible to elaborate and apply the model.
3.2.
Method and study description for the trial
The study used a combination of questionnaires (with Likert-scales and written answers) and debriefing interviews, as well as making video- and audio- recordings of the sessions for later analysis. The IPT systems used were a TAN VR-CUBE at Chalmers University in Gothenburg and a four-sided Trimension ReaCTor at University College London (for reasons of space, this article does not include full descriptions of hardware, software input/output devices, and network connections here, but see Schroeder et. al. 2001). The two systems were networked so that two people could collaborate with each other while they were in different places. There were six pairs – three pairs of strangers and three pairs of friends – who spent at least 210 minutes doing the five tasks together. The longest time that one pair spent together was 230 minutes. For one pair of friends, the trial was stopped after the second task because both partners experienced severe nausea and anxiety. The pairs of friends were all good friends who knew each other well. Subjects took a break of between 15 and 20 minutes between each task, and a longer lunch break of 60-90 minutes between the first two and the other three tasks. The times that pairs spent on each task were between a minimum of 30 minutes and a maximum of 70 minutes, with most tasks falling into the 40-60 minute range. All subjects first entered an open space with portals to the various task worlds described below. Before embarking on the task, they were given a chance to familiarize themselves briefly with using the system, and to
get acquainted with their partners. The environments they entered and tasks they carried out – in the order in which they were done - were as follows: Rubik’s cube puzzle – the task was to do a small-scale version of the popular Rubik’s cube puzzle with eight blocks with different colours on each side so that each side would have a single colour (ie. four squares of the same colour on each of the six sides).
Figure 1. The Rubik’s cube puzzle Landscape orientation – the environment in this case was a small townscape with surrounding countryside ringed by mountains. Subjects were instructed to familiarize themselves with this landscape and count the number of buildings. They were also told that they would be asked to draw a map of the environment at the end of the task.
Figure 2. The Landscape Whodo – again, this task was based on a popular game, in this case the ‘who-dunnit’ board game, Cluedo. The subjects were asked to find five murder weapons and five suspects in a building with nine rooms. They needed to locate the murder victim’s body and find and eliminate weapons and suspects.
Figure 3. Whodo Poster World – this environment consisted of a room with ten posters stuck on the walls. The posters each contained a list of six sentence fragments. When all the fragments were put in the right order, they would make a popular saying or phrase.
Figure 4. Poster World Modelling World – this environment contained 96 shapes (square blocks, cones, etc.) in six different colours. The subjects were told to make a building or model of a building to be entered into an architectural competition. They had to use at least three colours and the building had to be a single object. The result was to be their joint ‘architectural masterpiece’.
Figure 5. The Modelling World The tasks therefore ranged from an almost entirely verbal task with a fixed goal (Poster World) to an openended task that involved lots of navigation and manipulation of objects (Modelling World).
4. Applying the model of distraction-conflict theory to the networked Cave study To be able to make use of the model of distraction-conflict theory and make it suitable for analysing collaboration in shared virtual environments, the model was applied to observations from the networked Cave study described above (here after called the study).
4.1.
Focus on co-actor and task handling the technology while collaborating
Even if the experience reported on the Likert scale questionnaires showed quite similar responses from the subjects (Steed et.al. 2003), differences in the way they worked together could be observed from the video recordings where the pairs of friends differed in how focused on the task they were and the pairs of strangers how focused on each other they were (Spante et. al. 2003). Still, two pairs of strangers as well as the pairs of friends were able to solve the tasks as well as establishing and re-establishing their collaborative relationships over time. For these pairs their previous relationship didn’t seem to matter much. The pairs of friends had no major benefit over the pairs of strangers working together. Four of the five couples that finished the trial could coupe with the attention conflict between focus on co-actor and task in quite similar ways handling the technology. However, one pair of strangers differed quite a bit from the rest, not working well together and commenting negatively about working with each other. “I took an instant dislike of him since the very beginning. Maybe unfortunately” the London subject wrote in the questionnaire after the first task. “He was a point of reference, no support to the task” the subject in Gothenburg wrote. Also, in this pair one person had major difficulties using the technology during the trial day both with devises and applications. This could be observed on the videotapes as well as in the questionnaires.
Tabel 1. Subject ratings of being able to a) move around b) manipulate object c) find the way in the shared virtual environment, where 1=very difficult, 5=very easy (G is the person in Gothenburg, L is the person in London. G1L1 is the first pair, G2L2 the second etc. G1L1-G3L3 is strangers. G4L4, G5L5 is friends.) 5 4 3
q7a_task1
2
q7b_task1
1
q7c_task1
0 G 1 L1 G 2 L2 G 3 L3
q7a_task1 G 4 L4 G 5
L5
In the table above pair number 3 (G3L3), a pair of strangers, differs from the rest where they have a high difference in their rating of the first task. G3 finds it very difficult in contrast to his partner L3. It was also observed that they paid little attention to each other and could not establish a well functioning work relationships compared to the other couples that participated in the trial. In addition, the two of them differed in how focused on the task they were. One person was more eager to solve the tasks than the other (Spante et.al. 2003). This is an example where there was an attention conflict where they focused differently on each other as co-actors, as well as how they focused on the task. Moreover, the differences in how well they could handle the technology and the establishment of a relationship in the SVEs over time made an impact on the rather negative experience of working together and trying to solve the tasks.
4.2.
Focus on task and co-actor being distracted by technology
Above we could see how one pair differed from the rest in how focused they were on each other and on the task. We could also see how they differed in how they could handle the technology in comparison with the other pairs. However, in general the subjects disliked the second task out of the five they where supposed to do during the trial day. Five of the twelve subjects that did the task reported a slight nausea after this second task. During this task they were supposed to navigate around a large landscape, counting buildings and also memorizing the landscape in order to draw a map on a paper afterwards. Some answers of the question: “Do you feel physically uncomfortable, nauseous, or sick in any way after this task?” were the following: i) Yes-mild feeling of motion sickness, ii) a little head ache, iii) very much so yes, iiii) I feel a little tired in my eyes, v) a little dizzy. Another indication of discomfort were answers from some open-ended questions such as: i) It looked like a ghost town, ii) colours where dark and it was difficult to see the buildings, it was hard to “divide” the area. The environment makes it difficult to divide the labour, another subject wrote. Here the focus on the task made subjects run around a lot in the large environment in such a way and to such an extent that half of them reported that they felt some discomfort. At the same time they were focused on each other-indicating that the distraction came neither from the task or the co-actor, but rather from the structure of the environment in which the task was carried out i.e. the distraction came from the technology. It was after this second task one pair of friends had to stop participating in the trial due to severe nausea. In the debriefing interview one person said that even though he felt sick, it never crossed his mind to tell anyone. Also, when being inside the Cave doing the task and feeling sick, he wanted to take a rest laying down on the “grass” thinking: “nu går jag och lägger mig här på gräset” (now I’ll lay down on the grass).
5. An elaborated model of Distraction-Conflict theory When applying the model of distraction-conflict theory to the study of how people collaborated over an extended period of time, it becomes necessary to introduce some new aspects into the model. In particular, it has been shown that technology becomes important as well as time. Introducing both time and technology into the distraction conflict theory (Baron 1986), as presented by Hogg and Vaughan (1998:241), the model takes the following form:
Diagram 2. Elaborated Distraction-Conflict Theory Model Time Tendency to pay attention to co-actor
Individual performing task
Presence of co-actor
Social facilitation effects
Tendency to pay attention to technology
Attentional Increased arousal/
conflict
drive Tendency to pay attention to task
Social inhibition effects Time
In the new model we can see that the attentional conflict includes the technology in use. How that attentional conflict is coped with and the increased arousal/drive that follows leads either to social facilitation or social inhibition effects. The arrows back from the ‘social facilitation’ and ‘social inhibition’ to ‘individual performing task’ indicates the time aspect in the analytical model-in other words, this is an iterative process. We have seen above that the way subjects in the trial paid attention to each other, the task and the technology simultaneously had an impact on their collaboration and how they related to each other during the trial. Here it has only been possible to give a few examples of how this elaborated model can be applied to collaboration in SVEs. Further research is needed to establish the range of situations in SVEs this model is useful for, as well as its limitations.
6. References Baron R.S. (1986) Distraction/conflict theory: progress and problems. in Berkowitz, L (ed) Advances in Experimental Social Psychology. Vol 19. New York: Academic Press. rd
Berkowitz, L. (1988) A Survey of Social Psychology 3 Edition. NY: CBS College Publishing Biocca, F., Harms, C., (2002) Defining and Measuring Social Presence: Contribution to the Networked Minds Theory Measure, In Proceeding of Presence 2002. Porto: Portugal. Blascovich, J. (2002) Social Influence within Immersive VEs. In Schroeder (ed) (2002) The Social Life of Avatars. London: Springer. Cruz-Neira, C., Sandin, D.J., Defanti, T.A., Kenyon, R.V., Hart, J.C., (1992) The CAVE: audio visual experience automatic virtual environment. Communications of the ACM 35, 6, 64-72. Heldal, I. (2003) Usability of Collaborative Virtual Environments. Technical and Social Aspects. Department of Technology and Society, Gothenburg, Chalmers University of Technology. Hindmarsh, J., Fraser, M., Heath, C. Benford, S.J, Greenhalgh, C., (1998) Fragmented Interaction: Establishing Mutual Orientation in Virtual Environments. CSCW’98 :217-26. Hogg, M., Vaughan, G. (1998) Social Psychology. London: Prentice Hall. Schroeder, R. (ed) (2002) The Social Life of Avatars. London:Springer. Schroeder, R., Steed, A., Axelsson, A-S., Heldal, I., Abelin, A., Wideström, J., Nilsson, A., Slater, M., (2001) Collaborating in networked immersive space: as good as being there together? Computers & Graphics 25, 5, 781-788. Spante, M., Heldal, I., Steed, A., Axelsson, A-S, Schroeder, R. (2003) Strangers and Friends in Networked Immersive Environments: Virtual Spaces for Future Living. Home Oriented Informatics and Telematics (HOIT) 2003, Irvine, California. Steed, A., Spante, M., Heldal, I., Axelsson, A-S., Schroeder, R., (2003) Strangers and Friends in Caves: An Exploratory Study of Collaboration in Networked IPT Systems For Extended Periods of Time. ACM Siggraph Symposium on Interactive 3D Graphics 2003, Monterey, California.