The E-Collaboration Paradox: Evidence from Two Empirical Studies

1
The E-Collaboration ParadoxEvidence from Two
Empirical Studies

• Ned Kock

2
The e-collaboration paradox

• Media naturalness proposition
• The less face-to-face-like a medium is, the more
  obstacles exist for effective communication.
• Compensatory adaptation proposition
• Often individuals avoid using unnatural media,
  but, if they decide to use them, those
  individuals compensate for the communication
  obstacles posed by the media.

3
Study 1

• An Experimental Study of 20 Process Redesign
  Dyads
• Analysis Method Used Comparison of Means (ANOVA
  Mann-Whitney U Tests)

4
Participants

• The research study involved subjects with
  substantial hands-on experience in process
  redesign in the defense sector, recruited from
  management and engineering ranks of a large
  defense contractor.
• Their ages ranged from 23 to 60, with a mean age
  of 35. Fifty-nine percent of the subjects were
  males.

5
Treatment conditions

• The subjects were randomly assigned to dyads and
  to communication media conditions.
• Each dyad completed two similar process
  redesign-related tasks using different
  communication media for each task.
• Half of the dyads (i.e., 10 dyads) completed one
  of the tasks face-to-face while the other half
  completed the same task electronically.
• After this, all dyads moved on to the next task,
  using different media than they had used in the
  previous task that is, the dyads previously
  interacting face-to-face now interacted
  electronically and vice-versa.

6
Electronic communication medium

• A set of Web-based threaded online discussion
  boards created the electronic communication media
  employed in the experiment.
• The online discussion boards were developed using
  Microsoft Frontpage 2000 and Active Server Pages.
• They were used in a quasi-synchronous manner
  i.e., they were refreshed at short time
  intervals, creating an online chat-like
  environment where the discussions were threaded.
• One Web-based online discussion board was created
  for each dyad. All online discussion boards were
  identical.

7
Variables and measures

• Cognitive effort
• Measured based on NASAs task load index (a.k.a.
  NASA-TLX) developed by Hart and colleagues (Hart
  and Staveland, 1988).
• Communication ambiguity
• Measured based on an instrument previously
  developed and validated by Kock (2001), answered
  on a 1 to 7 Likert-type scale.
• Message preparation
• Measured based on an instrument previously
  developed and validated by Kock (2001), answered
  on a 1 to 7 Likert-type scale.

8
Variables and measures
9
Variables and measures

• Fluency
• Measured by counting the number of words
  exchanged by the members of the dyads and
  dividing it by the number of minutes each dyad
  took to complete the task (Kock, 1998).
• Task outcome quality
• Measured by comparing the process sketches
  generated by the dyad members with correct
  models (Kock and Murphy, 2001). Two different
  coders generated similarity scores used to
  assess task outcome quality independently.

10
Results
11
Results
12
Results

• The results suggest that the use of an
  e-collaboration tool, when compared with the
  face-to-face medium, increased perceived
  cognitive effort by about 41, perceived
  communication ambiguity by about 80, and
  perceived message preparation by about 47, while
  at the same time reducing fluency by
  approximately 77.
• The study also suggests that the use of the
  e-collaboration tool had no significant impact on
  the quality of the outcomes generated by the
  dyads.

13
Conclusion

• The study supports the e-collaboration paradox
  notion
• Media naturalness proposition
• The less face-to-face-like a medium is, the more
  obstacles exist for effective communication
  obsv. cognitive effort, communication ambiguity.
• Compensatory adaptation proposition
• Often individuals avoid using unnatural media,
  but, if they decide to use them, those
  individuals compensate for the communication
  obstacles posed by the media obsv. message
  preparation, fluency, task outcomes.

14
Study 2

• An Survey Study of 290 New Product Development
  Teams
• Analysis Method Used Structural Equation
  Modeling (PLS)

15
Participants

• Contact persons in a variety of technology-based
  companies in the Northeastern USA were selected
  to participate in the study.
• To be included in this study, each company must
  have developed a product that had been launched
  into the marketplace and commercialized for at
  least six months.
• Data from 290 new product development projects in
  66 companies were obtained.

16
Research instrument

• A questionnaire developed based on previous
  research on NPD teams (Kessler and Chakrabarti,
  1999 Lynn et al 2000) was used.
• All constructs in the study were measured using
  multiple-item scales, which in turn were
  Likert-type scales (0 Strongly Disagree to 10
  Strongly Agree).

17
Constructs and measures
18
Constructs and measures
19
Structural model
H1 (neutral)
H3 (positive)
H4 (positive)
H6 (positive)
H5 (positive)
H2 (neutral)
20
Results
H1 (ß-.009)
R2.226
H3 (ß.351)
H4 (ß.460)
R2.123
H6 (ß.488)
H5 (ß.090)
Notes T values were calculated through the
bootstrapping method Significant at p lt .001
in a one-tailed test
H2 (ß.102)
R2.315
21
Conclusion

• The study supports the e-collaboration paradox
  notion
• Media naturalness proposition
• The use of electronic communication (as opposed
  to the face-to-face medium) induces procedural
  structuring, which suggests an attempt (through
  procedural structuring) to compensate for media
  obstacles.
• Compensatory adaptation proposition
• Procedural structuring has a stronger effect on
  task outcome variables than electronic
  communication use, as individuals compensate (or
  perhaps overcompensate) for the communication
  obstacles posed by the electronic communication
  media.

22
Implications

• Development of e-collaboration tools in
  situations where compensatory adaptation is
• Desirable and unhindered
• Desirable but hindered (e.g., cognitive fatigue
  situations)
• Undesirable (e.g., entertainment-related B2C
  situations)

23
Key references
Final slide

• Kock, N. (2004), The Psychobiological Model
  Toward a New Theory of Computer-mediated
  Communication Based on Darwinian Evolution,
  Organization Science, V.15, No.3, pp. 327-348.
• Kock, N. (2001), Compensatory Adaptation to a
  Lean Medium An Action Research Investigation of
  Electronic Communication in Process Improvement
  Groups, IEEE Transactions on Professional
  Communication, V.44, No.4, pp. 267-285.
• Kock, N. (1998), Can Communication Medium
  Limitations Foster Better Group Outcomes? An
  Action Research Study, Information Management,
  V.34, No.5, pp. 295-305.

Available from http//www.tamiu.edu/nedkock/
24
Early development of symbolic communication
artifacts, e.g., cave paintings, about 30,000
years ago.
Emergence of modern humans, i.e., Homo sapiens
sapiens, about 100,000 years ago.
FtF communication through complex speech.
Emergence of early Homo sapiens grades, e.g.,
Homo sapiens neanderthalensis.
FtF communication through rudimentary speech.
Emergence of genus Homo, e.g., Homo habilis and
erectus.
FtF communication through facial expressions and
simple sounds.
Emergence of the genus Australopithecine, e.g.,
Australopithecus afarensis and africanus.
25
E-communication behavior theories
26
E-communication behavior theories (contd. 1)
27
E-communication behavior theories (contd. 2)
28
E-communication behavior theories (contd. 3)