Utility

1
Computer Supported Cooperative Work as Sub-field
in CHI

• Utility
• Importance of groups
• Importance of communications as an integral part
  of computing systems
• Interpersonal computing is a growth area in
  computer systems
• Groups are important, but not perfect
• Unaided groups don't live up to their potential
• Current technology constrains what groups can do
• Science
• Lewin Nothing is as practical as a good theory
• Reversed Nothing generates theory as a well as
  useful application
• Malone Challenge is to develop general theories
  of coordination that transcend type of actor
• Both goals require an interdisciplinary enterprise

2
Groups are valuable

• Way to pool resources to tackle problems that are
  too large or complex for an individual to solve
• Effort - e.g., construction gang, large software
  development projects
• Expertise - e.g., teaching this course, executive
  team
• Interests - e.g., school board, Congress
• Perspective/Point of view - e.g., human subjects
  review board
• In many task groups do better than the
  individuals comprising them
• E.g., Learning
• Interacting groups learn concepts more quickly
  and use different strategies (e.g focus)
• Students often learn better thru cooperative
  learning teams in schools than through
  individual instruction
• Mechanisms for why groups are better than
  individuals
• Aggregation of resources -- energy, ideas, points
  of view, etc
• Error checking
• Cognitive division of labor (e.g., in learning
  tasks, group is able to hold the hypotheses
  tried and their outcomes
• Synergy

3
What is CSCW

• Building information systems that help groups of
  people accomplish their goals
• Applying knowledge from
• Computer science
• Telecommunications
• Organizational behavior
• Small group research
• Individual cognition and motivation
• Task domains
• But the reference disciplines are inadequate to
  the task the practitioners don't look deeply
  enough
• Understanding the impact of information systems
  on the way groups work, play, live

4
What is CSCW (conceptual)
5
Core social science knowledge relevant to
CSCWdesign

• Much CSCW design is ad hoc, based on personal
  experiences limited observation (e.g.,
  contextual inquiry)
• This approach ignores a wealth of relevant core
  social science knowleged
• Small group tradition in social psychology
• Context-less group.
• IPO framework
• Small theories of relevant phenomena
• Not A THEORY of the group
• Social psychology tradition in organizational
  behavior
• Teams in organizational context
• What can the theory offer?
• Identifies leverage points
• Insight into design solutions
• But not a blueprint for design

6
Case 1 Social loafing

• People work less hard when they are working
  together than working alone or side-by-side
• Physical tasks
• Cognitive tasks

7
Karau Williams, 1997

• Subject perform a brainstorming task, working
  side-by-side
• Vary individual vs. collective work
• Individual Put ideas in separate boxes
• Collective Put ideas in common box
• Vary group cohesion
• Friends vs strangers
• Vary perceived ability of others in groups
• Low Im lousy at this type of task
• High Irrelevant comments or Im generally good
  at this type of task

Karau, S. Williams, K. The Effects of
Group Cohesiveness on Social Loafing and Social
Compensation Group Dynamics Theory, Research,
and Practice. June 1997 Vol. 1, No. 2, 156-168
8
Theory must account for these facts about social
loafing

• Social loafing reduced if
• Task is attractive
• Task is simple
• Group is attractive
• Individual's output is visible
• Expect others to perform poorly
• Own contribution is unique
• Research is done in field setting
• Individual is socialized to be altruistic (?)
• Women, eastern cultures, young children

9
Social loafing The theory

• Assumption that individuals work to the extent
  that they think their effort will lead to some
  valued outcome
• Being in a group may shift beliefs about the
  necessity of one's output, the efficacy of one's
  output, and the desirability of the outcome

10
Exercise

• On the Internet, people under-contribute
  resources to the groups they belong to
• Amazon reviews
• Posts to bboards on listservs (10/1 lurker to
  poster ratios)
• Free-riding on Napsters/Gnetella (70 share no
  songs top 1 of sites deliver 30 of song)
• Given what you know about social loafing, are
  there any design changes you can make that would
  lower the free-riding rates? Concretely, redesign
  http//www.sharedexperience.org/

11
Applying the theory to design

• Discussion

12
How would you handle

• Who can join
• Screen on domain expertise
• Prior relationships
• What can get talked about Moderation vs. lack of
  moderation
• How should people be identified Anonymity vs.
  aliases vs. real identities
• How large can the group get?
• Unlimited vs. capping vs. splitting

13
How would you exploit

• Identifiability?
• Attractiveness of task?
• Attractiveness of group?
• Group size?
• Uniqueness of contribution?
• Expectation that others will perform poorly?
• Self-selection?

14
Why isnt the theory more useful?

• Social science typical shows bivariate (or low
  order) relationships among variables, all else
  being equal
• Design demands understanding the complex of
  relationships
• Implications
• Need for intellectual tools for
  modeling/simulating complex systems. (But there
  is a problem is testing the models)
• Current social science provides inspiration, but
  no simple heuristic for translation to design

15
Inherent difficulties in applying group theory to
design

• CSCW tool is attempting to optimize multiple
  outcomes simultaneously
• Each desired outcome is multiply caused
• Each system feature may have effects on multiple
  psychological states and group processes

Psychological states/ Group process
Desired outcomes
System features
O1
F1
S1
S2
O2
F2