Delay and Trust in Distributed Collaboration

1
Delay and Trust inDistributed Collaboration
New Millennium, New WorkplaceImplications of
Changing Workplace Relationships

• Matthew Bietz
• School of Information
• University of Michigan

2
Overview

• Critical incident/Case study
• Distributed scientific collaboration
• Communication breakdown led to a loss of trust in
  the collaboration
• Implications for trust in distributed
  collaborations

3
Data Collection

• International collaborative HIV/AIDS research
• Teams from North America and Africa
• Interviews with scientists about collaboration
  practices
• 4 subjects independently brought up a recent
  communication breakdown in the lab
• Supported by the John Evans Foundation, the
  Waterford Project, and National Science
  Foundation grant IIS 0085951.

4
The Incident
Trust Breakdown

• African lab attempted a new assay that had been
  developed in North American lab
• This assay takes at least 4 weeks to complete
• Requires growing cell cultures
• After an on-site training period, Africans could
  contact remote expert for help

5
What Happened?
Trust Breakdown

• In week 3, African lab technician noticed
  something weird with the cell cultures
• Africa lab e-mailed the expert, but got no
  response for several days
• If the expert would have responded within 1-2
  days, the assay could have been salvaged

6
Consequences
Trust Breakdown

• Lost time and money
• Reputation of African lab was tarnished
• Africans lost trust in the reliability of the
  remote expert

7
Trust in Organizations

• Related to productivity, citizenship behaviors,
  and other metrics of organizational performance
• Jones George, 1998
• Allows individuals to take risks provides
  emotional stability
• Porter, Lawler, Hackman, 1975
• Provides a framework for interpreting other
  peoples behavior and intentions
• Dirks Ferrin, 2001

8
Trust in Distributed Contexts

• Trusting relationships are more difficult to form
  over distance
• Jarvenpaa Leidner, 1999 Zack McKenney, 1995
• Less common ground and social interaction
• Rocco, Finholt, Hofer, 2001
• Narrow information channels provide fewer social
  cues
• Bos, Olson, Gergle, Olson, Wright, 2002

9
Two Kinds of TrustMcAllister, 1995

• Cognitive Trust
• Expectations about competence and reliability
• Based on good reasons
• History, position, title
• Will colleagues do their jobs well and on time?

• Emotional Trust
• Provides interpersonal sensitivity and support
• Affective bond that develops among co-workers
• Do colleagues have good intentions?

10
Trust Breakdowns

• Earlier research suggests emotional trust is hurt
  more by distance than cognitive trust
• Rocco et al., 2001
• In this incident, we see that cognitive trust is
  also subject to breakdowns

11
After the failure in the lab

• Subjects still trusted the remote experts
• Intentions (emotional trust)
• Competence (cognitive trust)
• But no longer trusted the remote experts
  reliability (cognitive trust)
• We cant expect to get the help we need when we
  need it from the expert.

12
As a result...

• Subjects were less willing to take risks with the
  expert
• They demanded that the expert be on-site the next
  time the experiment was run
• But subjects still felt warmly toward the expert,
  and believed that the expert had good intentions

13
Discussion

• The same factors that increase risk in
  distributed work also make developing trust more
  difficult
• Reduced interpersonal interaction
• Narrow information channels
• Distance effects on cognitive trust
• Being reliable in person may be different from
  being reliable at a distance
• Strong relationships buffered the collaboration
  during trust breakdowns

14
Implications

• Managing trust deserves high priority in
  distributed projects
• Expensive to build, but can be quickly destroyed
• Trust is built on expectations they must be
  realistic (Deutsch, 1973)
• Distributed work requires different expectations
• Context for trust is extremely important
• Need to understand how distance changes
  assessments of risk and trustworthiness