2003 Lee Iverson leeiece ubc ca EECE 571

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Paper Reviews Goal: A paper is reviewed in order to – understand – situate – evaluate – in context of • a group (e. g. the class, research group) • a field (e. g. CSCW, …) • a task (e. g. building a system)

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Review Structure (possible) Goals – Why was the paper written? – What is it trying to demonstrate? Context – What field is it in? – What was the state of knowledge when it was written?

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Review Structure (cont’d) Summary – What does the author claim? – What hypotheses are tested or proposed? Analysis – Did the author succeed wrt. the goals? – Are the claims supported? – Are there things you didn’t understand? – Did you agree with authors conclusions?

© 2003, Lee Iverson <leei@ece. ubc. ca> Today • • 4 papers (+2) 5

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1984: Typology of Tasks Goals: • Provide a categorization of tasks performed in group settings that are: – mutually exclusive – exhaustive – logically related – useful

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1984: Typology of Tasks Context: • Social psychology • Body of work that had observed analysed task-oriented behaviour • Need to provide a means of organizing these findings to aid in understanding of task-oriented behaviours

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Typology of Tasks ing lann 1. P mp ve oti d-m 5. Co Cogn nfl ict itive 4. D 8. 7. C o xe Mi QII: Choose s on i s i ec o ch y s P 6. Conflict e or t o -m QIII: Negotiate etiti ve QIV: Execute 3. Inte llectiv ive t rea 2. C Cooperation QI: Generate

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1984: Typology of Tasks Analysis: • Useful model – Quadrants organized by processes – Subtypes make clear distinctions • Distinction between tasks that assume cooperation with tasks that recognize and resolve conflict is important.

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Suchman 1983: Office Procedure… Goals: • Provide work models that reflect actual practices • Provide framework for producing “office automation” systems

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Suchman 1983: Office Procedure… Context: • Social Anthropology • Office automation was focus of much development effort in ‘ 80 s Goal: Provide tools that would increase productivity by introducing computers to traditional offices • Existing work based on procedural models

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Suchman 1983: Office Procedure… Summary: • Identifies problems w/procedural model – unable to handle informal activity • Proposes practical action model – focus on meaning of actions – how actions contribute to goals, tasks and groups • “What are procedures for practitioners of office work? ”

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Suchman 1983: Office Procedure… Summary: • Observation of real workers on site • Analysis of conversations related to “Accounts Payable” – Problem to be solved – Outside of normal procedures – Characterize ways in which conversations serve the larger task

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Suchman 1983: Office Procedure… Findings: • Systems need to be designed so that communications and procedures can be modified to produce “smooth flow” in exceptional cases • Office automation is not a desirable goal • Systems should assist any work needed to reach goals

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Suchman 1983: Office Procedure… Analysis: • Place existing practice “under the microscope” • Probably better than designing systems to align users with restrictive assumptions of “best practices”

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups Goals: • Develop theory of task-oriented group activities • Explore consequences of theory – Analysis of patterns of behaviour – Implications for system designs

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups Context: • Sociology • Most theories of small group behaviour come from lab-based studies – Social psychology – Simple, artificial tasks ÞLimited generalisability • New emphasis on dynamics of groups

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups 1. Groups are complex social systems 1. Have relationships to (functions) 1. Organizations they are inside (production), 2. Their own members (member-support), and 3. The group itself (group well-being). 2. Have purpose in terms of shared goals 3. Partially nested 1. Complex membership relationships 4. Loosely coupled

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups 2. Group actions have modes: I. III. IV. 3. Inception (Goal choice) Technical solution (Means choice) Conflict resolution (Policy choice) Execution (Goal attainment) Modes are not fixed sequence, but kinds of activity to categorize particular actions of members

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups 4. Group behaviours show temporal patterns, including: 1. 2. 3. Flow of work Time-activity matching Entrainment or synchronization

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups 5. Collective action can be described by Temporal Problem Org. Response Ambiguity Schedules Individual Response Residual Problem Temporal committment Deadlines Conflicting Synchroniz requirements ation Norms for behaviour Norms for smooth flow Scarcity of time Comparts for interactions Regulation of interaction Resource allocations

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups 6. 7. Efficient workflow requires complex matching of activity bundles to periods of time Social entrainment is useful for constructing temporal patterns

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups 8. TIP Theory: Group interaction process refers to small scale flow of work in groups 9. TIP Theory: At any point, a group has a focal task 10. TIP Theory: Every action can be categorized as germane or not wrt. the current focal task

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups 11. Acts have situated (not generic) meaning wrt. modes, functions and paths of group activity. 12. Aspects of work flow are reflected in different ways of aggregating acts.

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Mc. Grath 1991: TIP: A Theory of Groups Analysis: 1. Seems like useful model 1. Emphasizes context and purpose of group activity 2. Flexible in a variety of situations 3. Does have some implications for how to think about design of systems

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Kraut 200 x: Applying Social Psych… Goals: 1. Suggest ways in which social psychology can inform research toward CSCW goals: 1. Support distributed groups 2. Enhance work of collocated groups 2. Introduce theory of “production loss” 3. Show knowledge can be applied to design of online groups

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Kraut 200 x: Applying Social Psych… Context: 1. Social psychology 2. Mixture of motivations from engineers/CS and social theorists 3. Build on work of Mc. Grath and others

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Kraut 200 x: Applying Social Psych… Summary: 1. Build on Input-Process-Output models 2. Recognize that outcomes sometimes conflict: 1. Star communication model leads to better problem-solving but reduces group satisfaction 2. Skeptics in brainstorming groups improve performance but reduce satisfaction

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Kraut 200 x: Applying Social Psych… Social loafing: “Group membership allows individuals to reduce their own effort towards group goals. ” 1. Cultural phenomenon 1. Asians, women and children 2. Western, men and adults 2. Varies with task type and group composition 1. Individually valued tasks 2. Lack of trust in group 3. Own unique contribution

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Kraut 200 x: Applying Social Psych… Production loss: Reasons groups don’t live up to aggregation effect 1. Social pressure 2. Social loafing 3. Production blocking

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Kraut 200 x: Applying Social Psych… System Design Suggestions: 1. Analyse tasks in terms of production loss 2. Categorize in terms of three reasons 3. Use strategies that combat reasons for loss Example: 1. Effects of anonymity on three reasons: 1. anonymity reduces social pressure 2. anonymity enables social loafing 3. anonymity irrelevant to production blocking

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Kraut 200 x: Applying Social Psych… Variable Link to Theory Design Identifiability • Direct deterrent to • Disallow anonymity loafing • Behaviour connected to outcomes Attractiveness of task • Increases value of individual outcomes • Provide interactivity • Define topic clearly • Don’t constrain content Attractiveness of group • Increases valance of group outcomes • Define topic clearly • Recruit membership based on external relationships

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Finholt & Sproull • Goals: – Compare “real” group with electronic groups (mailing lists) – Understand effect of DLs on organizational behaviour – Provide framework for evaluating group activity – Evaluate DLs in that context

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Finholt & Sproull • Context: – Organizational Behaviour – LANs uncommon in 1988 – Internet was largely built on Usenet and email – Electronic groups are seen to be having increasing influence on organizations

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Finholt & Sproull Summary: • Groups are more important than individuals within organizations • Assume that egroups should be considered as secondary preference for “natural” groupings • Observe that some egroups behave like “real” groups

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Finholt & Sproull Summary: • Restrict their interest to behaviours that only exist online • DLs used for variety of purposes: – social groups – required (organizational) groups – discretionary work groups

© 2003, Lee Iverson <leei@ece. ubc. ca> UBC Dept. of ECE Finholt & Sproull Summary: • Assume that all conversational acts can be categorized as: – Interaction – Influence attempts – Identity maintenance • Go through every message on DLs and classify them