The Use of Peripheral Social Awareness Tools in Collaborative Systems
Vichita Vathanophas & Michael B. Spring
School of Information Sciences, University of Pittsburgh
When people are working face-to-face or in physical proximity, they have access to information about one another. However, when people collaborate remotely using groupware, this information is generally unavailable. Awareness of the activity of other members in the collaboration team and of changes in the shared work material may be very important in collaborative work systems. The goal then is to enhance user performance in collaborative system by providing awareness information that is accurate and valuable. Specifically, this research examined how such a tool impacts the quality of the work effort and the communications between group members in the completion of a collaborative authoring task. The research has produced empirical evidence that an awareness tool negatively influences the quality of the work effort and number of communications. There is some suggestion from the research that the number of communications may mediate the relationship between the use of the tool and quality of group task. Future research on group awareness tools needs to evaluate fundamental communication processes and factors influencing performance in group work.
Keywords: awareness tool, awareness information, collaborative work, communication
In non-computer mediated groups, people working in close physical proximity have access to a large amount of information about one another. This includes such things as the presence or absence of members of the work group, what they are working on, who they are working with, how actively they are working, how they feel physically and emotionally, etc. This information is obtained directly through communication and indirectly through observation of shared artifacts (Beaudouin-Lafon, 1994). When people collaborate via computer mediation, this information and the opportunity to access it are diminished. Increasing the amount of information about the group availability in a computer mediated collaborative support system should increase the group’s ability to complete the task.
Being aware of other members of a team in a collaborative environment involves knowing both what people are doing and what is happening to the shared information space or artifact. We believe that "social awareness tools" can capture and represent this border or social periphery. The question is whether this tool enhanced social periphery will improve task performance in distributed shared information spaces?
This research outlines issues and opportunities related to the use of social awareness tools to improve collaboration by contributing to individual awareness of the state of the information store and the status of the individuals contributing to some effort. Collaboration tools may be classed taxonomically in a variety of ways. For purposes of this discussion, tools may support core or peripheral activity and they may support the task or the interaction. Thus we might consider the breakdown shown in Table 1 where the focus of this research is on tools related to peripheral support of interaction. This assumes an environment where the tools to support the core task and interaction are already in place.
Research on computer supported cooperative work can be separated into two main categories. The first research area is about how people work cooperatively (Rosenberg & Hutchison, 1994). The second is about the design, building, and testing of systems to support cooperative work. (Fitzpatrick, et al., 1995; Wilson, 1991).
In the social aspect, informal communication naturally occurs to allow participants to know about each other, and serve as the framework within which collaborative tasks are completed. Kraut, et al., (1990) found the frequency of communication between any two people was strongly related to their geographical proximity. Informal communication can be facilitated easily in a face-to-face environment. Tools are needed in a computer mediated collaboration system to support these functions. To provide awareness information, that is, information about the presence or absence of members of the work group, what they are working on, who they are working with, how actively they are working, how they feel physically and emotionally, etc., can build informal communication over distance and help diminish the lack of physical proximity. In addition, Dix (1994) developed the framework to show general groupware issues. When there is a shared artifact in the group work, one participant interacts with the artifact, the others perceive the consequences of the action. Dix (1994) called this observation by the others “feedthrough”. Also, In order to refer to the particular artifacts, participants will use various means, which called, “deixis”. From the perspective of the framework, feedthrough and deixis would be weak in electronic cooperation. One possible way to capture the feedthrough or deixis in collaborative work is by supporting awareness information for participants who have direct communication via a shared artifact. Strauss’s theory of actions (1993) also pointed out the presence, co-presence and awareness of others in locales, which is a bridge link between the social and the technical need for collaborative setting design, as an important feature in the collaborative work.
Effective work groups have always relied on peripheral awareness of each other’s activities, such as sound leaking under office doors and casual encounters in the hallway. A tool that performs these functions might be called an awareness tool -- “the condition of being aware: cognizance, consciousness, perception, sense” (Roget’s II, p.67). Awareness is “an understanding of the activities of others, which provides a context for your own activity” (Dourish & Bellotti, 1992, p.107). Since, awareness is knowledge about a dynamic environment, awareness must be maintained through perceptual information gathered from the environment. Awareness across distance has meaning that it can lead positively toward communications and interactions, and perhaps, most importantly, that it can contribute to a shared sense of community (Dourish & Bly, 1992). People must perceive certain types of awareness information in their environment, know how to interpret its meaning in their particular context, and realize how to apply that information (Greenberg & Johnson, 1997).
A research team in the School of Information Sciences at the University of Pittsburgh has developed a system for collaborative authoring called CASCADE. CASCADE stands for Computer Augmented Support Collaborative Authoring and Document Editing. This study assessed the effectiveness of various tools that was added to CASCADE to keep you up-to-date on what other members of your team are doing. The awareness tool in CASCADE has had various enhancements added to show such things as the level of user activity, the current focus of their efforts, and their disposition to engagement.
The study involved a methodology for examining a social awareness tool designed to improve task quality and communication in collaborative writing task. There are two groups of subjects. The first group used collaborative authoring system with a social awareness tool for collaborative writing task, while the second used collaborative authoring system without a social awareness tool to complete the collaborative writing task.
The study was conducted with students at the University of Pittsburgh. These individuals were rewarded in this study if they stayed on the study until the end. Students filled out a demographic data pre-questionnaire indicating their experience with various kinds of hardware and software systems, as well as their attitudes toward group work and toward computer-mediated group work. During a thirty-minute period, the subjects received training in the use of the collaborative authoring system along with a non-participatory demonstration of the system.
CASCADE was configured to become a tool for collaborative authoring. All materials were available on-line for every participant. A number of designated machines, on a Local Area Network, were available for subjects. In CASCADE, documents were available electronically and annotations were displayed inside documents as colorful labels. Subject can view document content and create new annotations.
Subjects used the collaborative authoring system on workstations with and without the social awareness tools. Each subject accessed the collaborative authoring system in the computer lab in the School of Information Sciences, University of Pittsburgh. One group of subject was presented the CASCADE system with a social awareness tool (figure 1), while the other was using CASCADE without a social awareness tool.
A social awareness tool had been created in conjunction with the CASCADE system that provide a representation of the following:
1. Activity information is defined as information about what others are doing or have done. Activity value will range using fuzziness presentation. Pictures of a user will be least fuzziness if the activity value is high or most fuzziness if activity value is low. The amount of fuzziness will be relative to all members in a team for a particular document.
2. Availability information is defined as information about who is around and available from the collaboration group. Availability information will be shown to users as a slide bar, which will range from the most available to the least available value relative for all members in a team, depending on the availability value calculated.
3. Commitment/Disposition Information is defined as information about how willing person is to do more, which includes how a person views the project (positively or negatively). Commitment/ Disposition value will range from a negative, zero, or positive number. Pictures of a user will be green if the disposition value is positive or red if negative. The amount of red/green will be relative to all members in a team for a particular document.
Sixty subjects were randomly assigned to one of twenty groups of 3 members (the experimenter simulated a fourth artificial member) for a collaborative writing project. Of the two categories of groups, members in the first category group (Group-A) worked on the task using the CASCADE system with the social awareness tools. The second category group (Group-B) used the CASCADE system without the social awareness tools. Each subject worked in different place and was given a series of tasks to finish in one day. The working day started as normal working day, which was about 3 hour in the morning (started at 9am) and 3 hour in the afternoon (ended at 4pm.) with lunch break from noon to 1 pm.
Once assigned to a group, subjects remained in the same group throughout the duration of the study. Each subject worked on the group task for six hours, which separated into two sessions to finish a brochure in session I and user guide in session II.
The study used online document access. Printing copies as the means for completion of tasks was disallowed. Direct, personal communications between members of a group were prohibited. Groups were allowed to use email, and communication tools in CASCADE or any other electronic communication tools.
The following procedure was used for each of the groups in this experiment:
1. A thirty-minute training session was conducted at the beginning of the first session. Members in Group-A were trained on the CASCADE system with social awareness tool, while members in Group-B were trained only on the CASCADE system.
2. All subjects were asked to fill out a pre-questionnaire to collect demographic data and attitudes towards team work.
3. Both groups were given documents in two directories available on the CASCADE system to complete the task.
4. All groups were asked to follow the same general steps:
· read and analyze the document individually
· work on activities given to individual in a given time
· work in group to complete brochure in CASCADE on session I
· work in group to complete user guide in CASCADE on session II
5. Individual was asked to fill out a cohesiveness post-questionnaire.
6. Individual was asked to fill out a post-questionnaire that asked about satisfaction with the awareness information tool and agreement with the final product.
3. Data Collection
Quality of final product was evaluated using evaluation criteria. Furthermore, number of communication within in each group were counted and analyzed. In addition, questionnaires attempted to elicit from participants their subjective experiences with the system in the study. Subjects rated their experiences with the social awareness tools when the experiment was finished.
The questionnaire was examined subject opinions on expected, delivered, missing, appropriate, and inappropriate system functions. Their performances would be analyzed to determine the expectations of participants in a social awareness tool and the extent to which those expectations were met.
The Analytical assessment was applied for quality evaluation in this experiment. To specify detailed component parts, helped an individual reader to assess the writing task in the standard set of criteria. The Analytical method identified particular components that each reader needs to look at for a fairly accurate assessment of an overall written ability. Two readers gave each of these areas an appropriate mark from 1 to 10 (1=Poor,…, 10=Excellent). The final grade, which was the quality of each group, was the total of the component scores.
The quality and number of communications in each group were measured during the experiment. The group means to be compared had been specified prior to data collection. Hence, a one-tailed T-test Procedure was used to investigate the differences among groups at the 0.05 significant levels. For survey results, mean value for satisfaction was reported.
randomly assigned to one of two conditions. The demographics were used to check the
distribution. There are
no significant different proportions between groups (at
p<0.05) using the Chi-square test with gender, student, writer, writing used computer,
and experience with collaborative writing system. In addition, the Mann-Whitney test was used
to determine if there was a statistically significant difference in the distribution
of age, education background, experience with computers, and life
assessment. No significant difference
Two people evaluated the Brochures (first product) and User Guides (second product). Pearson Correlation was used to measure the association between two evaluators. In terms of brochure (first product), the quality was significantly correlated for both evaluators (r=0.485, p=0.030). In addition, the quality of user guide (second product) was significantly associated for both evaluators at a = 0.05 significant levels (r=0.509, p=0.022).
experiment, the user guide was considered as the product to measure the
performance of group work. The
experiment assigned participants to produce the first product (brochure)
because participants needed to get used to the collaborative writing system,
other members in their team and the writing task.
For mean score ofr quality of user guide (the final document
produced), as predicted groups working without the social
awareness tools was higher (M = 133.9000), than it was for the groups
working with the tools (M = 101.6000).
The difference was statistically significance (t(18)= 2.868, p(1-tailed)
=.005). In this experiment, number of communication is the sum of send mail messages and number of comments
made by all members in a team. Number of communication in groups , and the
difference was marginally statistical
significant (t (18)= 1.907, p(1-tailed)=0.036).
veattitudes participantsing Tcorrectness of decisions madesingsing (t(18)= -
In addition, there are some evidences from the data that the amount of communication is a mediator between use of tool and quality. A significant negative correlation (r=-.410, p(1-tailed)=.073) was found between use of the tool and amount of communication. In addition, a significant negative correlation (r=-.560, p(1-tailed)=.005) was found between use of tool and quality of the product. A positive correlation which approached significance (r=.343, p(1-tailed)=.049) was found between amount of communication and quality.
Besides the use of awareness tool, there is another variable that might affect the team work. The amount of activity that occurred during the writing task for members in a team was recorded during the experiment. This process variable might affect the dependent variables in working as a team. Amount of activity was measured as the number of actions taken on a document, i.e., editing, commenting, etc; and amount of time each member spent working on a document. It is possible that the effect of the awareness tools on the dependent variables could be moderated by amount of activity. The data suggests that the effect of activity level on quality and the effect of activity level on amount of communication were stronger in the group with awareness tool than in the group without tool. The one-tailed t-test was used to determine the mean difference between the low and high activity levels in group with awareness tool. There was a statistically significant effect for activity level on quality (t=-1.899, p(1-tailed)=.047) and activity level on amount of communication (t=-2.823, p(one-tailed)=.011). Therefore, there is evidence that quality and amount of communication were influenced by the activity level in group with awareness tools.
The findings in this study about awareness tool are contrary to the theory in the previous
section. The study expected that group with awareness tools could produce
better quality for group task. In fact, the awareness tools had negative effect on their performance.
However, these results are
because they demonstrate that awareness information existed
in group task and perceived by the participants who had
the tools. In addition, the group with
awareness tools produced the quality difference influenced by the amount of activity. When the amount of activity was high in a
work group, they produced higher quality than those with low activity. The negative impact may result from the
incorrect drive to complex task for social facilitation or social
loafing. Facilitation is gained if the
dominant response is correct ; otherwise, the performance is harmed. The dominant response is likely to be the
correct one on well-learned or simple tasks ; accordingly, the performance is
facilitated. On the other hand, the dominant response is likely to be
incorrect and leads to debilitated performance on
complex or not well-learned tasks (Zajonc, 1965). In this study, the collaborative writing task was complex and the
content of the writing task was very technical. Additionally, to understand and learn the awareness information might have
increased the complexity of the overall task.
As a result, low performance in groups working with an awareness tool
might have been avoided if the tool had become second nature, and not a source
of cognitive overhead. from
ocial loafing is
defined as the finding that participants working together put out less effort
than that participants working alone (Latane et al., 1979). People loaf when their output cannot be
evaluated, and because evaluation is not possible when outputs are pooled, it
is this aspect of working together that leads to loafing. Therefore, the low quality from the data
might be affect from the social loafing.
amount of activity
performed by each member in a team was considered as a process variable that
impacted team performance in this study.
There was evidence that the activity level in groups with awareness tools influenced
quality and amount of
communication. Participants in group with awareness tool perceived the amount of activity from awareness tool
; ;therefore, they knew how much each member
worked in a document. This awareness
information had an impact on the quality they produced and amount of communication. When they realized that their teams had low
activity, they produced low quality and reported low satisfaction. On the other hand, when they perceived their
team worked more, they worked and produced higher quality and
satisfaction. There was difference in quality from the low and high
amount of activity in groups with
the awareness tool, while there was no
impact from amount of activity on
quality in groups without the awareness tool
because members in group without
awareness tool did not have information
about amount of activity during the
writing task. There were
only the communication
that existed in the writing process ; therefore, their quality and satisfaction were not
varied due to the level of activity.
Awareness information presented to participants correctly
still might enhance rather than debilitate
the quality. It may even be the case
that better instruction or training in the creative and functional uses within an extended time of the simpler awareness
tool used in this would lead to a
better performance in group task.
The redesign of awareness
tool should be included in
to the future
research. It appears that it was suboptimal in given awareness information to
participants. Participants perceived
awareness information from the tools and showed their interest in the awareness
tool and its implication. However, it
failed to create a good
presentation that would be satisfied participants using the
tools. The context of awareness
information is used to ensure that individual contributions are relevant to the
group’s activity as a whole, and to evaluate individual actions with respect to
group goals and progress (Zhang, et. al., 1999). Awareness support systems are designed to provide distributed
people with context information for collaborative activity (Chen &
Gellersen, 1999). In this study,
awareness tools continuously captured and transmitted context information in
real time, participants realized individual’s contribution, but easily got annoyed
by the presentation of the information.
The redesign for user
interface and added
build-in function for awareness
tool should enhance the use of the
tools. Shneiderman (1987) stated that
redesign of the human-computer interface can make a significant ly difference in user satisfaction. The email function should be built into the awareness tool and easily
used. The additional information users
requested – file a person is working on; current level of activity, etc. should
be incorporated. Finally, awareness tool should
also be built into the CASCADE window in order to reduce the number of windows on the
Additionally, the reduction in effort for social loafing;
for example, providing the participants’ output to others, and assigning the
evaluation criteria to the performance as an individual and a whole, and the enhancement in effort in
social facilitation by providing
the simple task for participants, might enhance
the performance of group work.
Moreover, instruction or training in a creative and functional use of technology
would be a very useful supplement to continued research.
This avenue of study helps to determine the impact of
awareness tools in a complex computer mediated environment. Much can be learned from highly abstracted
and controlled experiments; however, it is ourconviction that there
are more detailed experiments that endeavor
to isolate the impact of individual components in a larger functioning
system. As this study develops
confidence in the contributing factors, there are other questions such as:
What other input factors besides
awareness tools that influence
the performance of the group task or the group process?
· How does the explicit recording of activity change the way users work?
· At what point does peripheral information change from being informative to being distracting?
· How does the size of the group impact the importance of social awareness?
the field study be conducted with less possible
limitations in this study? In conclusion, this research found limited preliminary support for
awareness information in collaborative work.
While the researcher continues to believe that awareness information is
critical in collaborative work, it is not clear how to present and maintain awareness
information for the user.
However, the study now has at evaluating on fundamental
communication processes and factors influence the performance
in group work.
One conclusion that can be drawn from this study is that
human interaction is a complex phenomenon.
Accordingly, researchers must
make cautious alternatives
in studying interaction as to which variables will
be examined and incorporated into theoretical systems. This research was
experimented to suggest this potential, and by directing out consequences, in
which awareness tool was implemented,tools, which to be a suggested variable that may
be worthy of more general theoretical concern.
Overall, this study found limited support that awareness information
provides an incremental value in collaborative work.
The social awareness tool in group work studied here was in a prototype stage. Therefore, their net effect was mixed. Still,
it is a finding,
are suggest ions
that some of the benefits hypothesized earlier may well be realized when the
technologies are more fully developed.
Thus, further examinations of the hypotheses of this study should be
conducted with more mature technologies and in settings that have possibly
Beaudouin-Lafon, M. (1994). Beyond the workstation: Mediaspaces and augmented reality. In G. Cockton, S. W. Draper, & G. R. S. Weir (Eds.), Proceedings HCI’94, (pp. 9-18). Cambridge: University Press.
Chen, D., & Gellersen, H., (1999). Recognition and Reasoning in an Awareness Support System for Generation of Storyboard-like Views of Recent Activity. GROUP 99, Phoenix Arizona USA, 356-364.
Dix, A. (1994). Computer Supported Cooperative Work: A Framework. In D. Rosenberg, & C, Hutchison (Eds.), Design Issues in CSCW (pp. 9-26). London: Sprigner-Verlag.
Dourish, P. & Bellotti, V. (1992). Awareness and Coordination in Shared Workspaces. CSCW’92. Conference Proceedings on Computer Supported Cooperative Work, October 31-November 4, 1992, Toronto, Canada, 107-114.
Dourish, P., & Bly S. (1992). Portholes: Supporting Awareness in a Distributed Work Group. CHI'92. Conference Proceedings on Human Factors in Computing Systems, May 3-7, 1992, Monterey, CA, 541-547.
Fitzpatrick, G., Tolone, W.J., & Kaplan, S. M. (1995). Work, Locales and Distributed Social Worlds. In H. Marmolin, Y. Sundblad, & K. Schmidt (Eds.), Proceedings of the Fourth European Conference on Computer-Supported Cooperative Work (pp. 1-16), September 10-14 1995. Stockholm, Sweden: Kluwer Academic Publishers.
Greenberg, S., & Johnson, B. (1997). Studying Awareness in Contact Facilitation. Position paper for the ACM CHI’97 Workshop in Collaborative Systems, organized by Susan E. McDaniel and Tom Brinck, Atlanta, Georgia, March 22-27.
Kraut, R. E., Egido, C., & Galegher, J. (1990) Patterns of Contact and Communication in Scientific Research Collaboration. In J. Galegher, R. E. Kraut, & C. Egido, (Eds.), Intellectual Teamwork: Social Foundations of Cooperative Work (pp. 149-171). Hillsdale, NJ:Lawrence Erlbaum.
Latane, B., Williams, K., & Harkins, S. (1979). Many hands make light the work: The causes and consequences of social loafing. Journal of Personality and Social Psychology 37, 823-832.
Roget’s II The New Thesaurus Third Edition, Houghton Mifflin Company, p.67, Boston, MA, 1995.
Rosenberg, D., & Hutchison, C. (1994). Introduction. In D. Rosenberg, & C. Hutchison (Eds.), Design Issues in CSCW (pp. 1-8). London: Springer-Verlag.
Shneiderman, B. (1987). Designing the User Interface: Strategies for Effective Human-Computer Interaction. New York: Addison-Wesley Publishing Company.
Strauss, A. (1993). Continual Permutations of Action. New York, NY: Aldine De Gruyter.
Wilson, P. (1991). Computer Supported Cooperative Work. Oxford, UK: Intellect Books.
Zajonc, R. (1965). Social Facilitation. Science, 149, 269-274.
Zhang, M., Wu. L., Sun, L., Li, Y., & Yang, B. (1999). VCS: A Virtual Environment Support for Awareness and Collaboration. ACM Multimedia ‘ 99 (Part 2) 10/99, Orlando, FL, 163-165.