Spanning Time, Distance and Diversity with Technology A Program of Research

1
Spanning Time, Distance and Diversity with
TechnologyA Program of Research

• Laku Chidambaram
• Traci Carte
• Michael F. Price College of Business
• The University of Oklahoma
• Norman, OK 73072, USA

2
Agenda

1. Theory Development
2. Stream of Studies
3. Synopsis of Study 1 (focused on Quantitative
   Analysis)
4. Synopsis of Study 2 (focused on Qualitative
   Analysis)

3
1. Theory Development
4
Motivation

• Managing diverse teams is fast becoming one of
  the most pressing challenges facing modern
  organizations (Tsui Gutek, 1999)
• What role can/should technology play?

5
Prior Research

• Relational Demography (Tsui OReilly, 1989)
• Social Categorization Theory (Turner, 1987)
• Similarity/Attraction Paradigm (Byrne, 1971)

6
The Group Formation Process

• Many theorists (i.e., Gersick, 1989 McGrath
  1991) have suggested that groups alternate
  between focusing on
• Relational activities (i.e., group well-being,
  member support, relational development)
• Production activities (i.e., task performance,
  project deliverables, work outcomes)
• Some of Gersicks work suggests that relational
  development may take precedence in early stages
  and production in later stages of groups history.

7
Collaborative Technologies (CTs)
CAPABILITIES COLLABORATIVE TECHNOLOGIES COLLABORATIVE TECHNOLOGIES COLLABORATIVE TECHNOLOGIES COLLABORATIVE TECHNOLOGIES COLLABORATIVE TECHNOLOGIES
CAPABILITIES E-mail Groupware (e.g., Lotus Notes) Group Support Systems (e.g., GroupSystems) Desktop Conferencing (e.g., NetMeeting) Chat Rooms
REDUCTIVE CAPABILITIES REDUCTIVE CAPABILITIES REDUCTIVE CAPABILITIES REDUCTIVE CAPABILITIES REDUCTIVE CAPABILITIES REDUCTIVE CAPABILITIES
Visual Anonymity High High High Low (with Audio) None (with Video) High
Equality of Participation Moderate Moderate High Low High
Synchronous Interaction No No (in most cases) Yes (in most cases) Yes Yes
ADDITIVE CAPABILITIES ADDITIVE CAPABILITIES ADDITIVE CAPABILITIES ADDITIVE CAPABILITIES ADDITIVE CAPABILITIES ADDITIVE CAPABILITIES
Coordination Support No Yes Yes (in some cases) Yes No
Electronic Trail Yes Yes Yes Yes No
Enhanced Capabilities Image File Transmission Document Storage Retrieval Decision Support Features Audio- Video- Conferencing Instant one-on-one Messaging
8
Impact of CTs
CAPABILITIES EFFECTS IMPACT EARLY IMPACT LATER
Visual Anonymity Reduces salience of surface-level diversity High Lower
Equality of Participation Provides a level playing field minority opinions heard High Lower
Asynchronous Interaction Slow interactions reduced ability to coordinate etc. High Lower
Coordination Support Enables tracking of people, projects, and priorities Lower High
Electronic Trail Easy retrieval of comm. provides audit trail Lower High
Enhanced Capabilities Richer comm. (audio-/video-) task support (decision support) Lower High
9
Challenging Conventional Wisdom
Conventional wisdom and existing work on
technology support for collaborative work
suggests this for teams in general
Group makeup Early stages Late stages
Homogeneous Face-to-face interaction Collaborative technologies added
Diverse Collaborative technologies introduced Face-to-face interactions added
We propose that, depending on the degree of
diversity, the opposite may be more appropriate
10
Theoretical Model
Group Context
Relational Interaction
Collaborative Technologies
Relational Conflict
Diversity
Surface-level Diversity
Cohesion
Diversity Perceptions
Outcomes Performance Satisfaction
Deep-level Diversity
Time
11
2. Stream of Studies
12
Stream of Studies
Study Focus Design/Methods
1. Managing diversity with collaborative technology Lab experiment Quantitative
2. Coordination and collaboration Lab experiment Qualitative
3. Technology choice in diverse groups Field study lab Quantitative Qualitative
4. Creativity in diverse teams Lab experiment Quantitative Qualitative
13
3. Synopsis of Study 1(Quantitative Analyses
Focused on the Interactions and Performance of
Diverse Teams)
14
Research Question

• Do the effects of collaborative technologies
  differ over time between diverse and homogeneous
  teams?

15
Hypotheses

• Hypothesis 1 Over time, diverse teams based on
  surface-level attributes (H1a) and deep-level
  traits (H1b)supported by CT will become more
  cohesive compared to their collocated
  counterparts.
• Hypothesis 2 Over time, diverse teams based on
  surface-level attributes (H2a) and deep-level
  traits (H2b)supported by CT will experience less
  conflict compared to their collocated
  counterparts.
• Hypothesis 3 Over time, diverse teams based on
  surface-level attributes (H3a) and deep-level
  traits (H3b)supported by CT will perform better
  compared to their collocated counterparts.
• Hypothesis 4 Over time, diverse teamsbased on
  surface-level attributes (H4a) and deep-level
  traits (H4b)supported by CT will experience
  greater satisfaction compared to their collocated
  counterparts.

16
Research Design

• Conducted field experiment using students at OU,
  Salisbury State, and Michigan Tech
• 22 virtual teams collaborated on a semester-long
  database project and used Yahoo! Groups
  exclusively (for communication and task-related
  exchanges)
• 22 collocated teams collaborated on same project
  communicating primarily in face-to-face settings
  and using Yahoo! Groups for task-related
  exchanges
• Team assignments were made in each treatment so
  that half the teams were diverse and half
  homogeneous

17
Project Details

• Phase 1 Conceptual model (rough draft)
• Phase 2 Conceptual model (final version)
• Phase 3 Logical design (normalized)
• Phase 4 Implementation (queries, forms, reports)
• Phase 5 Debriefing

18
Data Collected

• A variety of perceived and actual demographic
  data
• Surveys administered after each phase of the
  deliverable capturing
• Perceived diversity (surface and deep)
• Relational conflict
• Cohesion
• Outcome satisfaction
• Grade assigned by course instructor used as an
  actual measure of performance

19
Sample Demographics
Variables Virtual (n105) Collocated (n105)
Mean (SD) Mean (SD)
Age (in years) 22.7 (4.60) 22.3 (3.80)
Work experience (part time in years) 3.98 (4.02) 4.00 (4.00)
Grade point average 3.15 (0.44) 3.03 (0.57)
Gender Male81 Female24 Male87 Female18
20
Survey Response Rates Actual (Percent)
Survey response after Virtual Teams Collocated Teams
1st deliverable 92 (88) 79 (75)
2nd deliverable 82 (78) 78 (74)
3rd deliverable 76 (72) 80 (76)
4th deliverable 79 (75) 86( 81)
5th deliverable 57 (54) 59 (56)
21
Measures
Construct Number of Items, Source a
Perceived surface-level diversity Harrison et al., 1998 How different are the members of your group in their age? How different are the members of your group in their ethnic background? .48
Perceived deep-level diversity Jehn et al., 1999 The values of all group members are similar My group, as a whole, has similar work values My group as a whole has similar goals Members of my group have strongly held beliefs about what is important within this group Members of my group have similar goals Members of my group agree on what is important to the group .90
Relational conflict Miranda Bostrom, 1993-4 Group members made negative remarks about other persons in the group Group conflict tended to be interpersonal in nature The conflict expressed by group members was targeted at particular person(s) in the group Members confronted each other on personal matters .60
Cohesion Seashore, 1954 Do you feel that you are really a part of this group? If you had the chance to do the same kind of work in another group, how would you feel about moving? How does this group compare to other student groups on each of the following points? The way people get along together The way people work together The way people help each other, .93
Satisfaction with outcomes Chidambaram, 1996 Overall, I was personally satisfied with my groups performance My group produced valuable results during this phase I think my groups deliverable is good Overall, the quality of my groups output this phase was high .93
22
Results of Repeated Measures ANOVA
Hypothesis 1 Over time, diverse teams based on surface-level attributes (H1a) and deep-level traits (H1b)supported by CT will become more cohesive compared to their collocated counterparts Significant three-way interaction for surface-level diversity (Pillais trace0.239, F2,362.512, p.049). Follow-up analysis suggests Supported Not supported
Hypothesis 2 Over time, diverse teams based on surface-level attributes (H2a) and deep-level traits (H2b)supported by CT will experience less conflict compared to their collocated counterparts. Significant three-way interaction for surface-level diversity (Roys largest root0.154 F2,372.841, p.071). Follow-up analysis suggests Supported Not supported
Hypothesis 3 Over time, diverse teams based on surface-level attributes (H3a) and deep-level traits (H3b)supported by CT will perform better compared to their collocated counterparts. Significant three-way interaction for deep-level diversity (Pillais trace0.377, F4,742.961, p.025). Follow-up analysis suggests Not supported Supported
Hypothesis 4 Over time, diverse teamsbased on surface-level attributes (H4a) and deep-level traits (H4b)supported by CT will experience greater satisfaction compared to their collocated counterparts. Significant three-way interaction for both surface-level diversity (Pillais trace0.387, F4,744.432, p.003) and deep-level diversity (Pillais trace0.313, F4,743.434, p.012). Follow-up analysis suggests Mixed Mixed
23
Results H1a
24
Results H2a
25
Results H3b
26
Results H4b
27
Key Findings
H1 The most diverse teams had the largest increases in cohesion from start to finish and outscored their collocated counterparts by the end. However, the most cohesive groups overall were the moderately diverse collocated groups, who started and ended with the highest levels of cohesion among any group.
H2 The biggest beneficiaries of the technology were the most diverse groups (whose level of conflict rose the least) compared to any other group. Interestingly, the opposite was true of the most diverse collocated groupsthey had the highest increases in conflict.
28
Key Findings (contd.)
H3 The moderately diverse groups supported by technology had the highest performance scores by the end of the study. Technology support had the least effect on groups with the least diversity.
H4 While the satisfaction of technology-supported groups increased over time, they did not exceed that of their collocated counterparts. In fact, in the case of both minimally and moderately diverse groups, levels of satisfaction were always higher in the collocated setting than in the virtual setting. However, in the moderately diverse condition, by the last session, satisfaction levels had dropped among collocated groups while it had increased for virtual teams.
29
Summary of Findings

• Diverse groups without technology support seemed
  to start off well (in contrast to the literature
  on diversity), but then encountered steep drop
  offs in cohesion, task-based conflict and outcome
  satisfaction (in line with the literature)
• In contrast, diverse groups with technology
  support started off poorly (in contrast to our
  expectations) but then gained ground, especially
  in terms of improvements in cohesion and outcome
  satisfaction (in line with our expectations)

30
Summary of Findings (contd.)

• Surprisingly, technology had little impact on the
  task performance of any groupall started out
  poorly, improved and then flattened out
• Also, remarkably similar profiles along most
  dimensions for homogeneous groups (with and
  without technology support)
• Overall, technology seemed to act as a brake for
  the dysfunctional processes of diverse teams,
  rather than as an accelerator of their inherent
  value

31
3. Synopsis of Study 2(Qualitative Analyses
Focused on Coordination in Virtual Teams)
32
Key Challenges for Virtual Project Teams

• Managing Time to Accomplish Task
• Establishing shared schedules
• Responding to deadlines
• Dealing with time differences
• Using Technology to Accomplish Task
• Dealing with technological constraints
• Learning to use tools
• Integrating with task processes

33
Coordination

• Technology coordination refers to the integration
  of the available technological tools with the
  task deliverables (Montoya-Weiss et al., 2001)
• Conflicting results from the literature, based on
  whether technology coordination is emergent
  (Malhotra et al., 2001) or imposed (Piccoli
  Ives, 2003)
• Temporal coordination refers to the
  synchronization of these task deliverables with
  member schedules and team deadlines (Sutanto et
  al., 2005)
• Again mixed results direct (Maznevski Chudoba,
  2000) vs. indirect effects (Massey et al., 2002)
  individual vs. group mechanisms (Sutanto et al.,
  2005) imposed vs. emergent rules

34
Coordination and Capabilities

• Coordination occurs through the two sets of
  capabilitiesreductive and additiveprovided by
  collaboration technologies (Herbsleb, 2002
  Brander et al., 2000)
• We suggest that technology coordination
  predominantly occurs through reductive
  capabilities, while temporal coordination
  predominantly occurs through additive
  capabilities of CTsan idea drawn from the
  Task-Technology Fit model (Zigurs Buckland,
  1998)

35
Intertwining Strands of Coordination
Reductive Capabilities (Structure)
Additive Capabilities (Structure)
Temporal Coordination (Content)
Technology Coordination (Content)
36
Main Thesis

• A single type of coordinationtermed
  monochordic coordinationis unlikely to provide
  the requisite means for a virtual team to succeed
  
• Both types of coordinationtermed dichordic
  coordinationrepresenting intertwining strands
  (over the life of the team) are likely to provide
  the means for success in such settings
• No coordination, termed non-chordic, refers to
  the absence of any significant coordinationi.e.,
  no (or little) coordination content in either
  technological capabilityand is likely to be the
  least successful approach of all

37
Operationalizing Constructs

• A total of about 5,000 messages
• Coding for coordination
• Two coders coded two teams independently used
  for training
• All differences discussed and resolved for
  consensus
• Remaining 20 teams were split between the two
  coders
• One more team was done together to check
  consistency inter-rater reliability was 93.5
• Counts for each coordination category split by
  session
• Coding for technology capabilities just completed
  (by two different coders)

38
Coding Protocol (Coordination)
CONSTRUCTS DEFINITIONS EXAMPLES
Technology Coordination How to use the technology to complete shared deliverables Coordination exchanges related to how the team will develop the database schema make changes divide sub-tasks who will post what, who will make changes etc. Ill post things in the files area for the professor to review. Let me set up an area for us to keep track of our draft work separate from our final deliverables.
Temporal Coordination How to stay on track and get the deliverables completed on time Expressions of structuring group activities which include references to scheduling, deadlines and turn taking Hopefully well have the whole group posting two or three times weekly in order to get this project done Id like to have an idea of what we are doing so I can play around this weekend.
39
Summary of Results

• All teams, regardless of coordination type,
  started offnot surprisinglyat about the same
  place in terms of their performance (as indicated
  by F2,19 .074, p.929)
• However, by the last session, significant
  performance differences emerged (F2,19 3.341,
  p.057)
• These differences were consistent with our
  expectation Teams that engaged in dichordic
  coordination outperformed those teams that
  engaged in non-chordic coordination and, those
  that engaged in monochordic coordination fell in
  between

40
Initial Results
Performance of Virtual Teams with Different Types
of Coordination (Session 1)
Coordination Types N Mean Std. Dev Std. Error Minimum Maximum
Non-chordic 9 .7333 .1090 .0363 .5500 .9000
Monochordic 4 .7500 .0408 .0204 .7000 .8000
Dichordic 9 .7500 .1061 .0354 .6000 .9000
Total 22 .7431 .0955 .0204 .5500 .9000
ANOVA Performance of Virtual Teams in Session
1
Sum of Squares Df Mean Square F Sig.
Between Groups .001 2 .001 .074 .929
Within Groups .190 19 .010
Total .191 21
41
Final Results
Performance of Virtual Teams with Different Types
of Coordination (Session 4)
Coordination Types N Mean Std. Dev Std. Error Minimum Maximum
Non-chordic 9 .8459 .0663 .0221 .6933 .9400
Monochordic 4 .9125 .0753 .0377 .8000 .9567
Dichordic 9 .9215 .0598 .0199 .8000 1.0000
Total 22 .8889 .0721 .0154 .6933 1.0000
ANOVA Performance of Virtual Teams in Session 4
Sum of Squares df Mean Square F Sig.
Between Groups .028 2 .014 3.341 .057
Within Groups .081 19 .004
Total .109 21
42
Coordination and Performance

43
Qualitative Results

• Analysis of dichordic coordination exchanges
  revealed ongoing discussions about
• how team members would use technology
• (e.g., Let me set up an area for us to keep
  track of our draft work separate from our final
  deliverables.)
• and
• when individual members would work offline
• (e.g., I will start compiling a list of
  attributes tonight and tomorrow and get back with
  you guys.)
• and
• when they would meet online
• (e.g., Lets all meet at 1000 p.m. to go over
  the attributes in our final ERD.)

44
Conclusions

• Embracing asynchroniety Teams that interwove
  temporal and technology coordination to stretch
  time performed the best
• Replicating familiarity In contrast, those that
  tried to overload timetypically by meeting
  together simultaneouslyperformed the worst
• One or the other Teams that relied on one form
  or another of coordination fell in between

45
Questions?