An Experimental Investigation of Information Sharing

1
An Experimental Investigation of Information
Sharing

• Daphne Raban
• Graduate School of Business
• University of Haifa
• draban_at_univ.haifa.ac.il

2
Presentation Outline

• Why study information sharing?
• Explanations for reduced sharing
• Theories
• Information ownership
• Research question and variables
• Hypotheses
• Experiment
• Results
• Discussion

3
Why Study Information Sharing?

• Integral part of work and business
• A significant application of IS
• Contrasting sharing behaviors

4
Explanations of Reduced Sharing

• Information commons invite the tragedy of the
  commons, free riding
• Database-mediated sharing (Connolly and Thorn,
  1990)
• Interpersonal sharing (Constant, Kiesler et al.,
  1994)
• Too much effort required
• Asymmetry

5
Theories

• Communality (Fulk and Flanagin, 1996 Wasco and
  Faraj, 2000)
• Pro-social transformation (Constant, Kiesler et
  al., 1994)
• Critical mass (Constant, Kiesler et al., 1994)
• Diffusion of responsibility (Barron and Yechiam,
  2002)
• Leadership (Butler, Sproull et al., in press)
• Perception of ownership (Jarvenpaa and Staples,
  2000 2001)

6
Information Ownership

• Uneven ownership may create free riding (Adar and
  Huberman, 2000), OTOH, when ownership is even is
  sharing needed?
• Ownership and self-enhancement (Heider, 1958
  Beggan, 1992)
• No transfer of rights
• Ownership is not clear an opportunity for
  framing by IS
• Is free riding all that bad in IS?

7
Research Question and Variables

• How can system-induced ownership status influence
  sharing?
• IV ownership status (3 levels)
• DV willingness to share (WTS)

8
Research Hypotheses

• H1 WTS for privately-owned expertise will be
  higher than WTS information depicted as an
  organizationally-owned document
• H2 WTS for a privately-owned document will be
  higher than WTS for information depicted as an
  organizationally-owned document

9
The Experiment
10
Results

• 173 participants, 3 treatments, each received one
  public and one private request for information.

11
Results of ?2 Tests

• H1 WTS for privately-owned expertise will be
  higher than WTS information depicted as an
  organizationally-owned document

12
Results of ?2 Tests

• H2 WTS for a privately-owned document will be
  higher than WTS for information depicted as an
  organizationally-owned document

13
Further Analysis

• A McNemar test showed a significant difference
  between WTS for personal and public requests for
  help (plt0.03)
• There was no significant difference for WTS of
  expertise and private document

14
Discussion

• WTS was significantly higher for expertise and
  private document than for organizational document
• WTS was significantly higher for private requests
  than for public requests
• A simple system-induced depiction of ownership
  created a significant difference in WTS
• Lack of difference between expertise and private
  document indicates ownership has more influence
  than information source

15
Discussion

• No problem in using the IM application a total
  of 262 personal message and 484 public messages
  were posted
• 13 of the personal messages were WTS
• 4 of the public messages were WTS
• Sharing induced more by personal appeal
• Public channel used more to broadcast than to
  interact
• Possible diffusion of responsibility in the
  public channel

16
Conclusion and Future Research

• A simple system-induced depiction of ownership
  created a significant difference in WTS
• Why are people more inclined to broadcast
  information publicly but share information more
  in private?
• Do people who participate more also help more?
• Is information sharing perceived as a solution to
  information overload or to information
  deficiency?
• What is the interplay between information
  sharing, searching and purchasing?

17
Thank you for listening!
draban_at_univ.haifa.ac.il