Group Style Differences Between Virtual and F2F Teams
Leonard Branson, University of Illinois at Springfield
Thomas S. Clausen, University of Illinois at Springfield
Chung-Hsein Sung, University of Illinois at Springfield
Face-to-face (F2F) teams form and function differently than computer-medi-
ated (virtual) teams. The social processes associated with effective team work are
different in F2F and virtual teams. These differences affect the ability of groups
of people to successfully form a team that can function effectively. This study
found that computer-mediated teams differ significantly from F2F teams along
important group style dimensions as measured by the Group Style Inventory
Keywords: teams, virtual, co-located
Introduction This study investigates the differ- pared to the F2F teams, where more
In recent years, there has been ences in group styles between virtual team-wide collective behaviors and
a growing interest in understanding and face-to-face (F2F) teams. We hy- information exchange were observed.
the behaviors of computer-medi- pothesize that virtual teams will form He concluded that greater team-wide
ated (virtual) teams and factors that and function differently than F2F collective behaviors gave rise to im-
impact their effectiveness. Due to teams. The group styles that form in proved information sharing activities
advancements in computer technol- virtual and F2F teams are evaluated among F2F team members.
ogy and the Internet, as well as the by using the Group Style Inventory Jarvenpaa and Leidner (1999) ex-
increasing need for collaboration be- (Cooke and Szumal 1994). This study amined the challenges of creating
tween and within companies, the use found that the type of team (virtual and maintaining trust in a global vir-
of virtual teams is on the rise. It is versus F2F) has a systematic effect on tual team. The authors reported on a
increasingly clear that computer-me- the group style. series of descriptive case studies on
diated communication systematically global virtual teams that worked on
and significantly impacts the social- Literature Review a common collaborative project with
ization processes necessary for effec- Andres (2006) studied the impact computer-mediated communication
tive teaming. Social and interper- of communication medium on vir- and whose members were separated
sonal cues that provide the basis for tual group processes. His study in- by location and culture. The authors
social intelligence are either absent vestigated the hypotheses that team conclude that trust can exist in teams
or are hard to detect in a virtual en- structure and communication mode built solely on an electronic network.
vironment. Having adequate social would impact the evolution of virtual Schmidt, Montoya-Weiss and
intelligence is a necessary, but not a group processes. The author studied Massey (2001) compare individuals,
sufficient condition for the establish- the behavior of software development F2F teams and virtual teams in the
ment and maintenance of healthy teams. The teams developed detailed area of new product development ef-
relationships. Working in a virtual design documentation for specified fectiveness. Two experiments exam-
environment tends to deny us the enhancements to a hypothetical uni- ined the effectiveness of new product
social cues that we use to interpret versity information system working in development project continuation
our social context, practice appropri- F2F or videoconference settings. The decisions. The first study compared
ate self-monitoring, and accurately research indicated the videoconfer- individual versus F2F decision-mak-
gauge the impact our actions are hav- ence teams exhibited more subgroup ing effectiveness. The second study
ing on others. information exchange when com- compared the decision-making effec-
Spring 2008 • Vol. 23, No. 1 65
Branson, Clausen and Sung
tiveness of individuals, F2F teams and Cooke and Szumal (1994) devel- der correlation and t-test statistical
virtual teams. They concluded that oped the Group Style Inventory and methods and reached the conclusion
teams make more effective decisions categorized group styles as construc- that virtual teams exhibit the distinc-
than individuals, and virtual teams tive, passive/defensive, or aggressive/ tive group interactive styles found in
made the most effective decisions. defensive. Constructive styles exist traditional F2F teams. However, few
Branson, Moe and Sung (2005) when team members are trying to studies have examined the differences
found that virtual teams process less satisfy their higher order needs (need in group styles found in virtual teams
information than individuals when for affiliation and achievement). The and F2F teams. This study examines
engaged in decision making. It ap- constructive style taps the full poten- the group style differences between
pears that F2F teams often use more tial of group members and produces F2F and virtual teams by using the
information and make better deci- effective solutions. It also enables Cooke and Lafferty Group Styles In-
sions than individuals, while virtual group members to fulfill both needs ventory (GSI) to measure and com-
teams use less information than in- for personal achievement as well as pare the structure and social processes
dividuals or F2F teams (Branson, needs for affiliation, allowing the full of virtual and F2F teams.
Sung, Decker, He 2005; Coopman potential of group members to be re- The GSI measures twelve dimen-
2001). Virtual teams spend more alized while facilitating effective so- sions of group styles, which collapse
time managing the team processes lutions by group consensus. Passive/ into the three group styles: construc-
and less time in processing informa- defensive groups behave in ways that tive, passive/defensive, and aggres-
tion and decision making, even when fulfill their security needs by placing sive/defensive. The central question of
the task is a decision-making task. greater emphasis on fulfillment of this study is “does team type (virtual
Branson et al. (2005) found that F2F affiliation goals only. They are inter- versus F2F) systematically and sig-
teams processed more information ested in maintaining harmony in the nificantly effect how teams form and
than individuals, and that individu- group, and accept limited information function, as measured by the Group
als processed more information than sharing, questioning and impartiality. Style Inventory?” Consequently our
virtual teams when making a perfor- The passive/defensive style team will hypotheses are:
mance appraisal decision. accept less than optimal solutions.
Alge, Wiethoff and Klein (2003) Team members will accept decisions H1: There will be no difference
examined whether temporality - the which have not benefited from con- in the group styles of virtual
extent to which teams have a past structive differing, creative thinking and F2F teams.
or expect to have a future together and individual initiative. Aggressive/
H1. A: F2F teams will be higher on
– affects F2F and virtual teams’ abil- defensive groups are characterized by
the constructive styles than the
ity to communicate effectively and competition, criticism, interruptions
make high quality decisions. For and overt impatience (Cooke and
teams lacking a history, results in- Lafferty 2003). The aggressive/de- H1. B: Virtual teams will be higher on
dicate that media exacerbates differ- fensive style emerges when members the passive defensive/styles
ences. F2F teams exhibited higher approach the problem in ways in- than the F2F teams.
openness and information exchange tended to help them maintain their
than virtual teams. status/position and fulfill their need H1. C: Virtual teams will be higher
Warkentin and Beranek (1999) for security by task related activities. on the aggressive/defensive
discussed the effect of communica- Aggressive/defensive groups are con- styles than the F2F teams.
tion training on virtual group in- cerned with their need for power and
teractions, especially for enhancing need for control. Social intelligence literature indi-
rational links and thereby improv- The majority of Cook and Szum- cates that multiple cues are necessary
ing communication and informa- al’s research was based on F2F teams. for accurate social perception, and
tion exchange in virtual teams. They Potter and Balthazard (2002) ex- these cues have a critical influence
concluded that teams that were given panded Cook and Szumal’s work on explicit categorization judgments
appropriate training exhibited im- by investigating virtual teams using (Kubota and Ito 2007). The social
proved perceptions of the interaction group styles inventory techniques to perception and cognitive psychology
process over time, specifically with determine whether factors that drive literature indicate that once a catego-
regard to trust, commitment and conventional team performance exist rization judgment is formed, it di-
frank expression. in virtual teams. The authors exam- rects future perceptions, information
Successful teaming requires effec- ined the decision process and per- gathering, information storage, and
tive socialization processes. In order formance styles of 42 virtual teams decision-making (Krzystofiak, Cardy,
to assess the socialization processes, and analyzed the data using zero-or- Newman 1988). Relying exclusively
66 Spring 2008 • Vol. 23, No. 1
Branson, Clausen and Sung
on text messages denies team members Not surprisingly, teams will form employed 11.4; 64.5 percent were fe-
the additional cues of voice, facial ex- and function differently based on the male; 24 percent were post bachelors
pression, body language and personal assigned task. Because task type can level, with 4.6 percent holding ter-
demeanor. Similarly, the virtual team systematically affect the form and the minal degrees (PhD, JD, MD). The
environment is not good at conveying function of a team, this study con- subjects were mostly night students
confirmatory communications like a trolled the effect of task type by hold- pursuing professional development.
smile to accompany a sarcastic com- ing the task constant. Holding the All the subjects had extensive expe-
rience in team work, and most used
virtual teaming in their current jobs.
“…F2F teams should be more likely to form There were no significant differences
in the demographics of the two types
constructive teams than virtual teams.” of teams.
ment. The computer mediated team is task constant allows us to more fully Results
not well suited for conveying unam- understand the effect of team type In order to determine if there is a
biguous messages and verifying mean- (F2F versus virtual) on how the teams systematic style difference between
ing through other communication formed and functioned. All teams in the two groups (virtual and F2F
cues (Haythornthwaite and Neilson this study were assigned an intellec- teams), we conducted a t-test on the
2007). Virtual teams that are depen- tual/analytical task. In this study, each twelve group style indicators of the
dent on computer mediated commu- team member was provided with a GSI for the thirty-two virtual and the
nication have fewer social cues to use business case with financial and non- thirty F2F teams. The results are pre-
in social intelligence gathering, build- financial information related to the sented in table 1.
ing trust, close friendships, or complex performance of eight organizational The means are significantly differ-
relationships. F2F teams have more units. Each team member was asked ent between virtual and F2F teams
social cues to use in social intelligence to prepare an individual performance in the constructive style group; the
processes, and should be more likely appraisal of the manager of each of self-actualizing, humanistic-encour-
to establish trust, and productive so- the eight organizational units. Subse- aging, and the affiliative dimensions.
cial relationships. Consequently, F2F quently, each team was asked to pre- The mean scores for constructive
teams should be more likely to form pare a team performance appraisal of style: self-actualizing, humanistic-
constructive teams than virtual teams. each unit manager. Members of each encouraging, and affiliative dimen-
While all teams can have character- team had the same data, but the data sions were significantly (at the .05
istics of all three types of teams, they was systematically manipulated be- level) higher for the F2F teams. The
will tend to have a dominant style. tween teams to investigate the type achievement oriented dimension was
and amount of information included higher for the F2F team, but the dif-
Methodology in the decision process of individuals, ference was not statistically signifi-
Task type can have a systematic F2F teams and virtual teams (Bran- cant. While virtual teams did score
effect on team formation and func- son et al. 2005). Each team completed higher on the passive/defensive style,
tioning. Driskell, Radtke, and Salas the GSI to investigate whether teams the differences in the scores were not
(2003) developed a task classification differed systematically based on team statistically significant. All the other
system which used six basic catego- type (F2F versus virtual). styles and dimensions did not have
ries: (1) mechanical/technical tasks, The GSI was administered to six- significantly different means (at the
requiring the construction or opera- ty-two teams of students (with 3-4 .05 level).
tion of things; (2) intellectual/analytic members) at a major Midwestern uni- An analysis of covariance was con-
tasks, requiring generation of ideas, versity. Thirty of the teams were F2F ducted to see if team type (virtual or
reasoning, or problem solving; (3) teams and thirty-two of the teams F2F) or any of the demographic vari-
imaginative/aesthetic tasks, requir- were virtual teams. The virtual teams ables had a significant effect on the
ing creativity or artistic endeavor; (4) did all their work using Blackboard group style. Results indicate that team
social tasks, requiring training, sup- and other virtual tools such as e-mail, type had a significant (at the .05 level)
porting, or assisting others; (5) ma- fax, and telephone. The members of impact on the achievement-oriented,
nipulative/persuasive tasks, requiring the virtual teams were on different humanistic-encouraging and affilia-
motivation or persuasion of others; campuses of the university, did not tive dimensions of the constructive
and (6) logical/precision tasks, requir- know each other, and were not able to style. Team age (average age of team
ing performance of routine, detailed, meet in person. The average age of the members) had a significant (at the
or standardized tasks. subjects was 31.6 years; average years .05 level) effect on the “conventional”
Spring 2008 • Vol. 23, No. 1 67
Branson, Clausen and Sung
dimension, which tends to have high
pressure to conform and team mem- Table 1
bers who think alike, avoid innovative Group Style Inventory Differences Between Virtual and Face-to-Face Teams
or creative ideas, and make poor de-
cisions. Team gender (all male, all fe-
male, or mixed gender) had no signifi- Mean ( ± Standard Error)
cant effect on any of the twelve group Style/Dimensions P-Value
dimensions. All other demographic CONSTRUCTIVE STYLE 76.46 (±1.6767) 70.20 (±1.5196) 0.0008*
variables had no significant effect on
Achievement-oriented (11) 18.73 (±0.4465) 16.51 (±0.4439) 0.3597
Self-actualizing (12) 17.79 (±0.5705) 17.12 (±0.4677) 0.0074*
Conclusions Humanistic-Encouraging (1) 19.55 (±0.4453) 18.02 (±0.4580) 0.0203*
Our results indicate that F2F Affiliative (2) 20.39 (±0.4008) 18.55 (±0.3982) 0.0018*
teams have significantly higher scores PASSIVE/DEFENSIVE STYLE 15.58 (±0.9493) 16.21 (±1.1696) 0.6807
on the constructive style. F2F teams Approval oriented (3) 3.74 (±0.3787) 3.10 (±0.2933) 0.1833
scored significantly higher on:
Conventional (4) 6.12 (±0.2625) 5.74 (±0.2971) 0.3477
1. Self actuating members are op- Dependent (5) 2.95 (±0.3618) 4.03 (±0.5242) 0.0977
timistic and interested, will offer Avoidance (6) 2.77 (±0.4036) 3.30 (±0.3971) 0.3555
any idea without worry, will en-
AGGRESSIVE/DEFENSIVE 9.55 (±0.7650) 9.14 (±1.3702) 0.7954
courage new ideas and unusual
perspectives. Solutions tend to be Oppositional (7) 3.30 (±0.2886) 2.76 (±0.4138) 0.2874
innovative, creative and of high Power-oriented (8) 2.21 (±0.2630) 3.14 (±0.4837) 0.0998
quality. Competitive (9) 1.19 (±0.1725) 1.01 (±0.2900) 0.6045
2. Humanistic-encouraging mem- Perfectionist (10) 2.85 (±0.2780) 2.23 (±0.4180) 0.2256
bers are sensitive, supportive of
other members; are generally * indicates significance at 0.05 level in means of F2F and virtual teams.
constructive; are interested in the
growth and development of fellow
group members; provide each oth- formation in a decision as individuals ing in a state of deindividualization,
er support and assistance; are able or F2F teams. Virtual teams appear to reduced evaluation concern, and a
to build on the suggestions/ideas form in a way that makes good group more impersonal and task-oriented
of other team members; and reach decision making difficult. As a result, attentional focus. Virtual team mem-
high quality decisions they are more concerned about issues bers interact more superficially and
other than making good decisions. For share fewer cues with team members,
3. Affiliative emphasis is placed on example, teams make suboptimal de- both of which make it more difficult
interpersonal relationships; mem- cisions when team members are more for them to develop an interest in
bers treat each other well, com- concerned about maintaining position the growth and development of team
municate openly and like to work and power than pooling information members, to build on the ideas and
together. Solutions are not always and developing more comprehensive suggestions of fellow team members,
the best, but team members sup- models of the problem that can then and to build positive constructive rela-
port the team decisions. activate new knowledge. tionships with team members. Virtual
Virtual groups are physically iso- team members have a much harder
Virtual teams scored higher on the lated and visually anonymous. Mem- time building a trusting relationship
passive/defensive, dependent, avoid- bers of virtual teams have a reduced with team members. Not only is com-
ance, and power-oriented dimen- set of social cues with which to con- munication less complete and satisfy-
sions, where group processes prevent trol the effect of stereotyping on their ing for virtual teams, but the building
effective teaming and lead to inferior relationship management. Research of trusting relationships is more infre-
decisions. Virtual teams are less able conducted by Benbunan-Fich, Hiltz, quent. Without good communication
to minimize the negative effects of and Turoff (2002) supports the idea and trust, it is difficult for any team
teaming on good decision making. that anonymity increases uninhibited, to function effectively. Virtual teams
Branson, Moe and Sung (2005) and hostile behavior and extreme decision have to be conscious of the problems
Branson et al. (2005) found that vir- making. These effects are due to the associated with the lack of communi-
tual teams do not process as much in- absence of interpersonal cues, result- cation in a virtual team, and members
68 Spring 2008 • Vol. 23, No. 1
Branson, Clausen and Sung
have to make a special effort to com- formation and often resort to behav- Branson, L., C. Sung, J. Decker, and F. He.
municate effectively. Virtual teams ing in less than socially intelligent 2005. An empirical investigation of the
can form trusting relationships, but it ways. As a result, it appears that virtual role of accounting outcome information
takes extra effort and skill to do so. team members tend to behave in ways and collocated and virtual team process-
Methods to enhance team interac- that are more consistent with pas- es on decision making. Paper presented
tion should be developed. The missing sive/defensive or aggressive/defensive at American Academy of Accounting
social cues from F2F interaction may behaviors. These behaviors can have and Finance, 10 December, in St. Pete
be included. The internet has provid- a significant impact on team perfor- Beach, Florida.
ed enhanced communication through mance on a decision-making task. A Cooke, R. and J. C. Lafferty. 2003. Group
the use of smiley faces (J) and other limitation of this study is the result of Style Inventory. Chicago: Human Syn-
emoticons. While these are not the using only intellectual/analytic tasks. ergistics, Inc.
same as the visual cues present in F2F The results are not generalizable to Cooke, R. A. and J. L. Szumal. 1994. The
interaction, the emoticons can give other types of tasks. impact of group interactions styles on
clues to an individual’s meaning and This research provides empiri- problem-solving effectiveness. Journal of
intentions as he interacts in a virtual cal evidence that for virtual teams to Applied Behavioral Science 30: 415-437.
environment. These additions to vir- be as effective as F2F teams, people Coopman, S. J. 2001. Democracy, perfor-
tual interaction take additional effort who work on virtual teams will have mance, and outcomes in interdisciplin-
on the sender’s part, but may enhance to learn more about the limitations ary health care teams. Journal of Business
the communication. Additional ways and problems with virtual teaming, Communication 38: 261-284.
to enhance virtual communication and develop effective strategies to Driskell, J. E., P. H. Radtke, and E. Salas.
should be investigated. overcome these limitations. Effec- 2003. Virtual teams: Effects of technol-
Our conclusion is that virtual teams tive communication that allows team ogy mediation on team performance.
have problems of both form and func- members to collect important social Group Dynamics: Theory, Research, Prac-
tion that are the result of their “vir- intelligence is an important part of tice 7: 297-323.
tual” nature. The virtual teams in this the solution. Technology has evolved Haythornthwaite, C. and A. Nielsen. 2007.
study scored significantly lower on faster than human sociology. We are “Revisiting Computer-Mediated Com-
the constructive style than the F2F on a new frontier of human relation- munication for Work, Community and
teams. The consequences of this find- ships, and it will take time to learn Learning: Intrapersonal, Interpersonal
ing are that virtual teams are less able how to successfully relate to and work and Transpersonal Implications.” In
to achieve the positive results possible with other people in these new and Psychology and the Internet. 2nd ed., ed.
in group decision-making perfor- rapidly changing technological envi- Jayne Gackenbach, 167-185. Elsevier
mance. Our results indicate that F2F ronments. Inc.
teams score higher on the self-actual- Jarvenpaa, S. and D. E. Leidner. 1999. Com-
izing style, which has members who Reference munication and trust in global virtual
are involved in the process, welcome Alge, B., C. Wiethoff, and H. Klein. 2003. teams. Organization Science 10: 791-
creative and innovative solutions and When does the medium matter? Knowl- 815.
support other members’ ideas. F2F edge-building experiences and opportu- Kaboto, J.T. and T. A. Ito. Multiple cues in
teams score higher on the human- nities in decision-making teams. Orga- social perception: The time course of
istic-encouraging style, which has nizational Behavior and Human Decision processing race and facial expression.
group members who are interested Processes 91: 26-37. Journal of Experimental Social Society
in each others growth and develop- Andres, H. P. 2006. The impact of commu- 43(5): 738-752.
ment, provide one another with assis- nication medium on virtual team group Krzystofiak, F., R. Cardy, and J. Newman.
tance and support, and constructively process. Information Resources Manage- 1988. Implicit personality and perfor-
build on each other’s suggestions. F2F ment Journal 19: 1-17. mance appraisal: The influence of trait
teams also score higher on the affilia- Benbunun-Fich, R., S. Hiltz, and M. Turoff. inferences on evaluations of behavior.
tive style, where members are strongly 2002. A comparative content analysis of Journal of Applied Psychology 73(3): 515-
committed to the group while focus- face-to-face vs. asynchronous group de- 521.
ing on reaching the best solutions. cision making. Decision Support Systems Potter, E. R. and P.A. Balthazard. 2002. Vir-
On the other hand, virtual teams 34: 457-469. tual team interaction styles: Assessment
score higher on group styles that lead Branson, L., B. Moe, and C. Sung. 2005. An and effects. International Journal of Hu-
to less optimal decisions. Even when empirical investigation of information man-Computer Studies 56: 423-443.
virtual team members have a highly processing in individuals and virtual Schmidt, J., M. M. Montoya-Weiss, and A.
developed social intelligence, they are teams. The Journal of Business and Be- P. Massey. 2001. New product develop-
unable to collect important social in- havioral Science 13(1): 234-242. ment decision-making effectiveness:
Spring 2008 • Vol. 23, No. 1 69
Branson, Clausen and Sung
Comparing individuals, face-to-face
teams, and virtual teams. Decision Sci-
ences 32: 1-26.
Warkentin, M.. and P. M. Beranek. 1999.
Training to improve virtual team com-
munication. Info Systems Journal 9: 271-
About the Authors
Leonard Branson is Professor and Chair
of Accountancy at the University of Il-
linois at Springfield. One of his streams
of research focuses on decision making
by individuals and groups.
Thomas S. Clausen is Assistant Profes-
sor of Accountancy at the University
of Illinois at Springfield. His research
focuses on Accounting Information Sys-
tems and Managerial Accounting topics.
Chung-Hsien Sung is Associate Profes-
sor of Mathematical Sciences at the Uni-
versity of Illinois at Springfield. He has a
wide range of research interests.
All sponsor schools are AACSB accredited
70 Spring 2008 • Vol. 23, No. 1