Activity patterns in intellectual collaboration

1
Activity patterns in intellectual collaboration
A Representation of Practice Knowledge in
Information Behavior
CSCW 2002 Workshop 5 Peter Jones, Redesign
Research, Dayton 16 November, 2002
2
Research Question

• How do scientists seek and use information in the
  conduct of research?
• Focus on
• Research with context of Program/Project
• Collaboration in research efforts
• Individual and social information use
• Analysis within across disciplines
• How is knowledge used in practice?

3
Background

• Elsevier Science sponsored study to understand
• Context of research work
• Information use by scientists
• Preparation of articles use of journals
• Approach
• Current Review and Analysis
• Design of Field Study
• First Field Site Case Western RNA Center

4
Research Methodology

• Rapid Ethnography (Millen, 2000)
• Research Task Notebook
• Ethnographic interview, observing, analysis
• Individual group walkthrough sessions
• Grounded theory approach
• Glaser, Strauss Process theories
• Iterative theory build test
• Field data applied to both product research
• Activity Theory Case Analysis
• Framework allows additive construction
• Supports multiple cases research database

5
Analysis and Representation Approaches

• Activity theory (Engeström)
• Social cognition, analysis of multiple activities
• Motivation, internal external, organization
• Focus on technology as instrument of work
  practice
• Contextual Design
• CD Models useful to triangulate reps
• Physical, Info Flow, Cultural/Organizational,
  Artifact
• Lifecycle concept
• Different temporal lifecycles of information use
• Borgman (1996) only ref to general info lifecycle

6
Analysis of 25 key studies
Context Cognitive / Analytical Social / Collaborative
1. Scientific Discipline 4 2
2. Institution or Department 2 1
3. Research Program or Project 1 (Review study) None
4. Individual 13 3
Levels in Activity Theory Activity (3), Action
(4), Operation Unit of analysis focused on
Research Project Activity levelWhere the work
of science gets done.
7
Field Research Project

• RNA Center for Molecular Biology
• 5 PIs and 10 Researchers
• Focus on info behavior in research
• Used diaries, interviews, observation, artifacts
  analysis, walkthroughs
• Information use motivated by knowledge production
  required for research
• Experiments, Articles, Grants, Exploration

8
Research Project Focus of research
work Projects can be studied, are
self-contained Instruments include most research
tools content of interest More defined work
roles, project activity similar within discipline
9
Research Project LifecycleExperimental Study
Molecular Biologist example
Info Use Score
10 9 8 7 6 5 4 3 2 1
Prepare article
Prepare Proposal
Validate findingsReferences lit
Review publishCycle
Exhaustive review
Develop issues
2
Admin wait
3
1
Develop findings,Compare studies
Identify key issues
Project start,Collect data
Initial Findings Proposal Start
Data Findings Validation Publishpaper
(Conference) Plan Collection Articles
10
(No Transcript)
11
(No Transcript)
12
Wheres the Collaboration?

• All PIs and most PhDs collaborate with other
  researchers (in this field)
• Not evident from observation
• Motivated by knowledge gaps
• Required expertise for research
• Specific equipment needed for exps
• Desire to work with best researchers
• Mediated by
• Conferences and specific meetings
• Email, phone calls, shared artifacts

13
(Initial) Information Flow Model
New data every day
Experiments
PhD StudentResearcher
Finding latest in the area
Research Articles
Review of DataStructuring Paper
Specific contributions
Collaborator
Principal Investigator
Discussion of Findings
Paper ideas, shared findings
14
Addressing Analysis Issues

• Data overload?
• Inductive analysis drawing from the field
  details means leaving stuff out
• Losing focus across findings?
• Grounded theory approach
• Coding transcripts for themes
• Iterating analysis, co-analysts
• Bridging design -
• Walkthroughs, field focus groups, analyzing data
  across independent studies

15
Addressing Representation Issues

• Iterative abstraction
• Not modeling further than the data shows
• Moving up but keeping it real
• Multiple models Beyond CD
• A good lesson from CD Use multiple
  visualizations to build a complete story
• Not being constrained by CD try
• Activity models, Temporal cycles, Social
  networks
• Glued together with case details