John Stanton
IRLS 575 Human Factors
Paper One
October 3, 2004
 
Three Papers relating to Computer Supported Collaborative Work

In this paper I review three research papers that were presented at the 2002
ACM conference on Computer Supported Cooperative Work. The first paper,
“Why distance matters: effects on cooperation, persuasion and deception” [1]
talks about the sociological effects of being in distant cities when participating
in Group Work situations. The second paper, “What is chat doing in the
workplace?” [2] is about using a modified group chat program to aid in CSCW.
And the third paper, “Social navigation: Making web sites be places for social
interaction” [3] presents the results of an experiment that created two web based
CSCW systems.

The below three sections contain some of the major points of the papers reviewed
as they relate to our class readings on CSCW. The material is extracted directly
from the paper in many cases and paraphrased in others. I selected information
from the papers that related to the lecture notes and readings from class.



Why distance matters: effects on cooperation, persuasion and deception

This paper examined how geographic distance affects collaboration using computer-mediated communication technology. It investigated experimentally the effects of cooperating partners being in the same or distant city on three behaviors: cooperation, persuasion, and deception using video conferencing and instant messaging (IM).
Even though this paper looks at the difference between participants in the same city ve. a distant city, it really relates only to the synchronous-remote case. This application, like the other two papers, deals with communications between users.

The following tasks were performed by participants in the experiment: the desert survival task, the Paulhus Deception Scales (PDS) and the prisoner’s dilemma game. The desert survival task tests the extent to which a participant can be persuaded to change her evaluation of the relative importance of items needed to survive in the desert. The PDS is an instrument used to examine scores for impression management and deception. The prisoner’s dilemma game (which is discussed in the Grudin reading and the lecture notes) is used to assess the willingness to cooperate. Together, these tasks measure the effects of technology and geography on persuasion, on the tendency for deception, and on cooperation.

Both the Video conferencing and instant messaging were performed in two rooms next to each other in Irvine, California. Both of the experiments tested synchronous communications. Half of the participants were told they were communicating with someone in the same city and half were told they were interacting with someone across the country in Boston. The participants believed that the person they were communicating with (the confederate) was located where they were told (i.e. Irvine or Boston). The confederate was a member of the research team who pretended to be located in Boston in half the cases and somewhere in Irvine in the rest of the cases.

It is important to note that neither effects of task order nor gender were detected in any of the analyses. Also, wrap up interviews strongly suggest that the confederate came across as ‘nice’ and ‘neutral’ with regard to her evaluation of the participants.

The results for the Paulhus Deception test showed a significant difference in deception by geographic location of the confederate. People who believed that the confederate was in a distant city gave more deceptive answers to her than those who believed that the confederate was in the same city.

Those in the same city condition were more persuaded by the confederate than those in the Distant city condition.

The Prisoner’s Dilemma game had 6 rounds. Participants who believed they were interacting with a person in a distant city were less likely to cooperate on the first two trials than those who believed the other person was in the same city. Cooperation then increased on the last two trials.

The confederate appeared the same in the video image and with IM wrote the same messages to all subjects. The amount of network delay was no different between the Same and Distant city conditions. Yet people regarded the confederate differently. They considered her different enough to change how they cooperated, deceived, and were persuaded by her. These effects cannot be attributed to the media, or technical differences, but rather due to social perceptions.

 
What is chat doing in the workplace?
 
This paper presents an empirical study of a synchronous messaging application with group-oriented functionality designed to support teams in the workplace. In particular, the tool supports group chat windows that allow members of a group to communicate with text that persists for about a day. An analysis of use shows that the group functionality was used primarily for bursts of synchronous conversations and occasional asynchronous exchanges. The content was primarily focused on work tasks, and negotiating availability, with some non-work topics and humor. This paper dealt with an application that was both synchronous-remote and asynchronous-remote.


This application was an example of taking a popular application (chat) and modifying it to have groupware extensions. The application was designed to provide unobtrusive accessibility.
An awareness of what one’s distant colleagues are doing, and their availability for interaction, are key parts of improving multi-site work. E-mail, the form of computer-mediated communication in widest use today, has only limited potential for signaling awareness and availability. E-mail is generally used as an asynchronous communication medium.

On the heels of tremendous popularity among recreational users, synchronous messaging applications are beginning to show up at work. Synchronous messaging in the workplace has a number of uses, including opportunistic interactions, broadcasting of information or questions, and a “signaling” function in which people negotiate availability for other interactions.

Rear View Mirror (RVM) is an implementation of an IM and presence awareness system with novel features designed to support teams. The identity of other users is shown by means of an iconized picture. Each person’s presence state is indicated by the border color around the picture, and more detail can be obtained with a mouse-over. Users can changes their status explicitly by selecting from a status menu, or by setting screen-saver-like timeouts that change status after a user-settable period of mouse and keyboard inactivity.

Groups in RVM are also associated with group chat windows that have persistent content. This provides a mechanism for teams to communicate with all other team members, and a way for a person to clearly separate communications associated with different teams. Each time a group member logs in, the group chat window opens, displaying (by default) the last days’ chat messages

Any RVM user can create any number of groups. By default, the creator becomes the group’s administrator. Groups can either be permissive (anyone can join) or the group administrator can restrict membership to a specific list of use.
New messages increment a counter displayed as the icon (in the upper left corner) of the chat window. Even when a chat window is minimized, a user can see the number of new messages since the last time the chat window was on top. In addition, certain events (people logging on and off, new private chats, and new messages) also can play sounds which can be turned on and off by the user.
13% of the messages concerned negotiating availability. This reinforces the view that chat is sometime used as a lightweight tool for seeking out and arranging heavier-weight interactions such as telephone calls and meetings. This was not, however, a very frequent use of the tool. There was a relatively small amount (8%) of conversation on non-work topics. The largest category of activity was actually doing work (69%). There were no incidents of “flaming,” or even any clear expressions of anger.

The content analysis shows that chat was used overwhelmingly for work discussions or for articulation work to coordinate projects and meetings, and to negotiate availability.

It is not clear if this pattern is typical of all synchronous messaging, however. Compared to IM, group chat is relatively public. When conversing one on one, there may be a greater temptation to gossip, flame, etc., since there is a much higher expectation of privacy. In addition, RVM associated each message with the user’s identity in unmistakable fashion, and there was no straightforward way to hide one’s identity, or create a false identity.

Some users who were familiar with IM reported that group chat tended to be less intrusive than IM. When an IM window pops up on one’s screen, it draws one’s attention, and people often feel obliged to respond. Group chat is different, in that the window is always there when one is logged on, but since many messages appear that are directed toward a group or other individuals, no one feels compelled to look at new messages immediately if it is inconvenient to do so.
 


Social navigation: Making web sites be places for social interaction

The paper presents a case study of the design of two social interaction Web sites: CHIplace and Portkey. Both Web sites described in the paper had the goal of fostering interactions among participants. The first Web site, CHIplace, was developed for the ACM CHI 2002 conference to extend the interaction opportunities and the interactions among people both in time and space. The second site, Portkey, was developed for the summer interns at IBM TJ Watson Research Center to enable the interns to exchange helpful information and experiences and to develop the social networks — personal and professional — that they needed to function effectively in their new environment.

Both of the web sites use asynchronous-remote communications even though Portkey took place at one location. Thw web sites were an example of taking a popular application (web pages) and modifying it to have groupware extensions. The application was designed to provide unobtrusive accessibility.
The designs of the two sites were driven by the need to identify the key sociological challenges, of which there were three:
Aside from the three key sociological challenges, the design of the two social interaction Web sites had two additional design challenges.
Both sites provided a number of mechanisms that enabled participants to construct and evolve a persistent and verifiable identity. Profiles contained real-life information about each person including their name, photograph, home country, email address, and a link to a Web page with more information. Both web sites provided a number of social browsing tools that enabled the participants to get to know the other participants of the site.

The authors created awareness about the sites to drive traffic to them. Invitations to participate were sent to various email distributions and groups. The authors also registered the CHIplace site with various search engines and placed links at ACM Web sites. To sustain awareness, participants were apprised of new additions and ongoing developments on the site.

Both sites had an incentive-based participation reward program. In one case there was a monthly $1000 prize for participating in the site given to one random participant.

There has been extensive CSCW research on supporting interaction among geographically distributed co-workers. The authors distilled the research results into a set of requirements and four basic social interface components that can foster social interactions:
The use of dynamically generated pages from templates and a database simplified adding and modifying pages and creating custom information for signed-in users.

A supportive infrastructure alone is insufficient to attract and retain user interests. New and frequent content additions, be it editorial or user contributions, must occur or else user interest disappears quickly.

Portkey was more successful in fostering discussions because, in part, the Portkey members had more interests and problems in common. Lightweight social interaction tools such as quick polls provide lower barriers to participation and produces valuable user contributions.

Information about other people, discussions, and community content is of great interest to the users. Making people and their contributions visible is invaluable in capturing visitor interest.






References:

1. Managing communications: Why distance matters: effects on cooperation, persuasion and deception
Erin Bradner, Gloria Mark
November 2002                 
Proceedings of the 2002 ACM conference on CSCW
 http://portal.acm.org.ezproxy.library.arizona.edu/citation.cfm?doid=587078.587110
 

2. I M everywhere: What is chat doing in the workplace?
Mark Handel, James D. Herbsleb
November 2002                 
Proceedings of the 2002 ACM conference on CSCW
http://portal.acm.org.ezproxy.library.arizona.edu/citation.cfm?doid=587078.587080
 

3. Social navigation: Making web sites be places for social interaction
Andreas Girgensohn, Alison Lee
November 2002                 
Proceedings of the 2002 ACM conference on CSCW
http://portal.acm.org.ezproxy.library.arizona.edu/citation.cfm?doid=587078.587098