Modeling the characteristics of virtual teams' structure Daphna Shwarts-Asher * Tel Aviv University, Tel Aviv, 69978, Israel E-mail:
[email protected] *Corresponding author
Niv Ahituv Tel Aviv University, Tel Aviv, 69978, Israel E-mail:
[email protected]
Dalia Etzion Tel Aviv University, Tel Aviv, 69978, Israel E-mail:
[email protected] Abstract: The main research question is whether output of virtual teams is affected by the teams' structure. A model, suggesting that the team structure can be manipulated in order to increase the team's output will be presented. The study used an experiment, in which subjects, who were grouped into teams, had to share information in order to complete an intellective task. The findings indicated that virtual teams output were inferior to face-to-face teams output, and that structure did overcome two negative impact of the virtuallity - less effectiveness and less satisfaction; yet, it did not raise the virtual team efficiency. Keywords: Management, e-infrastructure, e-Social Science, Virtual teams, Team structure, Team performance. Reference to this paper should be made as follows: Shwarts-Asher, D., Ahituv, N. and Etzion, D. (2009) ‘Modeling the characteristics of virtual teams structure’, Int. J. Business Intelligence and Data Mining, Vol. X, No. X, pp.XXXX. Biographical notes: Daphna Shwarts-Asher is a Ph.D. student at The Faculty of Management, Tel Aviv University, Israel. Her research interests are OB and economic and behavioral aspects of the Internet. The objective of the Ph.D. research is to investigate corporate output of virtual teams, assuming team structure can compensate for the shortcomings of virtuallity. She has presented the research at The 3rd International e-Social Science Conference (Michigan, USA) and at The 3rd International Conference on Complex, Intelligent and Software Intensive Systems (Fukuoka, Japan). In addition, she teaches courses in the field of management at several institutions. Niv Ahituv is the Academic Director of Netvision Institute of Internet Studies and the Marko and Lucie Chaoul Chair for Research in Information Evaluation at Tel Aviv University. From 1999 to 2002 he served as Vice President and Director General (CEO) of Tel Aviv University. From 1989 to 1994 he served as the Dean of Graduate School of Business Administration at Tel Aviv University. In 2005 he was awarded a Life Time Achievement Award by ILLA, The Israeli Association
for Information Technology. Professor Ahituv represents the Israeli Government in UNESCO in issues related to Information Technology. Dalia Etzion (PhD in Psychology 1975) has been with the Faculty of Management, Tel Aviv University since its establishment. She was one of the founders of the Organisational Behaviour Programme and headed the programme for many years. Prof. Etzion is a certified Psychologist and authorised trainer of Psychologists and Supervisors in the areas of Social, Industrial and Organisational Psychology. Her main interests over the years have been: Functioning of groups in organisations; consulting and organisation development; gender-differences in management and technology; job stress, burnout and life/work integration; respite and recovery. She has published extensively in highly rated professional journals.
1. Introduction There has been a transformation from individual work to team work in the last few decades (Ilgen, 1999), and many organizations use teams for many activities done by individuals in the past (Boyett & Conn, 1992). The use of virtual teams has also become common particularly in international organizations and global companies. In light of this growing phenomenon, the traditional definition of "team structure" and "team dynamics" should be redefined, as part of a revised model that predicts the influence of the virtuallity and structural levels on processes, social and tasks, that effect team output. This research contributes to a better understanding of virtual teams in hope of improving the teams work in the virtual world.
2. Literature review Investigation of team structure in the virtual environment holds significant promise for research and practice (Powell, Piccoli & Ives, 2004). Integration between virtual teams and structural characteristics raises the question "Has the virtual era put an end to team structure?" Leavitt (1996) claims that the rapid changes impose organizations to reduce structures. Hackman (2002), on the other hand, predicts that team structure will always exist and managers will continue to be bothered by team design. This section reviews virtuallity levels and structural characteristics of teams, in order to grasp an integrated view of the subject. Thereafter it refers to the group inter-dynamics, and will present Bales' IPA (Bales,1950), a method for analyzing the systems of human interaction.
Virtuallity Level Virtuallity level of a team has become an integral part of a team's definition (Martins, Gilson & Maynard, 2004). Many variables are affected by the virtual level of a team. Face-to-face team member are more cohesive (Huang et al., 2003), have stronger social ties (Warkentin, Sayeed & Hightower, 1997), are more dedicated to the task and to other team members (Olson & Teasley, 1996), have a stronger team identity (Bouas & Arrow, 1996) and have more affection to other team members (Weisband & Atwater, 1999), than in virtual teams. Strong social ties in virtual teams can be achieved but will take longer time than in face-to-face teams (Burke & Chidambaram, 1996). Many researchers have attempted to find the reasons why virtuallity has a negative influence on team output: frequency and distance (Cramton & Webber, 1999), the fact that team members are not familiar with one another (Gruenfeld et al., 1996), the difficulty in sharing information, and insufficient and confusing discussions (Thompson & Coovert, 2003). Another group of researchers compared communication technologies, assuming that technology limits information (Straus & McGrath, 1994). The comparisons concluded that face-to-face teams are more efficient than teams using video (Andres, 2002), and video communication is more efficient than audio (Burke et al., 2001), adding text into video or
audio communication improves performance (Baker, 2002), and satisfaction (Olson, Olson & Meader, 1997). Maruping & Agarwal (2004) show that teams tend to use different sorts of communication technologies for different kinds of interpersonal interactions.
Characteristics Team Structures Division of Labor The division of labor does not have a direct influence on team performance, but has an indirect influence, by means of perceived efficiency and team coordination (Strijbos et al., 2004), and is a stronger predication variable than individual characteristics (Ahuja, Galletta & Carley, 2003). The use of expertise assists in reducing errors (Potter & Balthazard, 2002a) and function diversity is important in achieving team efficiency (Bunderson & Sutcliffe, 2002). A team that is structurally diverse is one that its members have different positions or tasks and are distributed in different branches, can be exposed to unique information. In this manner, sharing unique external information elevates performance (Cummings, 2004).
Hierarchy In order to explain the importance of the manager in integrative groups, Maier (1967) compares the group to a starfish and the leader of the group to the starfish's central nervous system. When individuals act as an organized unit, they become a higher type of organization - a single whole organism. Even when there is no formal division of labor, the role of the leader is divided between team members (Johnson et al., 2002). When there is a formal leader, status labels has a strong effect on team members (Weisband, Schneider & Connolly, 1995). But a series of studies has shown that more importantly, effective virtual leadership is dependent on the ability to communicate (Kayworth & Leidner, 2001/2002 ; Tyran, Tyran & Shepherd, 2003 ; Kim, Hiltz & Turoff, 2002). In practice, leaders send more messages (and longer ones) than others (Yoo & Alavi, 2004).
Work Process In large teams meeting procedures are better known and there is better technology coordination then in small groups (Bradner, Mark & Hertel, 2005). Studies show that successful teams focus: on the task (Hofner, 1996), on structured goals (Huang et al., 2003) and on the development of routine (DeSanctis, Wright & Jiang, 2001). Successful teams also take a lot of time to understand the process and contents of the work (Iacono & Weisband, 1997), especially in the initial stages (Hause et al., 2001). Not only does the medium limit the team's ability to coordinate information (Graetz et al., 1998), virtual teams spend a great deal of time understanding how to execute the task (Lebie, Rhoades & McGrath, 1996), and team meeting that it's members are distant physically take longer time than face-to-face meetings (Siegel et al., 1986). Coordination process are related positively to performance and satisfaction (Powell, Piccoli & Ives, 2004), and become more significant as time passes (Burke & Chidambaram, 1995).
Social and Task Processes Hackman (1987) suggested that team design influences the team's output through its impact on processes. Chidambaram & Tung (2005) claimed that performance is driven by social demands and task demands. Bales (1950) developed a method to measure the team inter-dynamics, called Interaction Process Analysis (IPA). Bales' IPA is a method for analyzing the "systems of human interaction" in, originally, small face-to-face groups. The heart of the method is a technique of classifying behaviors act by act, as they occurs, and a series of ways to analyze the data and obtain descriptive indices of group process. The IPA consists of 12 complementary-paired group processes; these are further subdivided into four major functions, describing communications issues or problems, as shown in Table 1.
Function Social-Emotional Area: Positive Reaction Task Area: Attempted Answers Task Area: Questions
Process 1. Shows solidarity: raises other's status, gives help, rewards 2. Shows tension release: jokes, laughs, shows satisfaction 3. Agrees: shows passive acceptance, understands, concurs, complies 4. Gives suggestion: supplies direction, implying autonomy for other 5. Gives opinion: supplies evaluation, analysis, expresses feeling, wish 6. Gives orientation: supplies information, repeats, clarifies, confirms 7. Asks for orientation: askes for information repetition, confirmation 8. Asks for opinion: askes for evaluation, analysis, expression of feeling 9. Asks for suggestion: askes for direction, possible action Social-Emotional 10. Disagrees: shows passive rejection, formality, withholds help Area: Negative 11. Shows tension: asks for help, withdraws out of field Reactions 12. Shows antagonism: deflates other's status, defends/asserts self Table 1. Categories of Interaction Process Analysis (Bales, 1950)
3. In-depth interviews with virtual team managers and members Rather than building a theoretical model based only on the literature review, we decided to confront previous theories with the organizations' reality environment using in-depth interviews with virtual team managers and members. The purpose of the interviews was to deepen the understanding of virtual teams work in real life. The interviewees included employees from diverse organizations, who participate in virtual teams in order to complete different types of projects. They answered open questions in order to understand the organization, the nature of the team and its activities, how the team operates and the difficulties the team faces. The aim of the interviews was to focus on difficulties they face, by examining the development and existence of structural characteristics. A Summary of twenty interviews shows that virtual teams are a new phenomenon, in which managers lack the proper tools to deal with. It is obvious to virtual team members that virtual teams operate without structural characteristics. The form of communication (virtual compared to face-to-face) will determine whether the processes will be task or socially oriented: virtual communications is task oriented, while face-toface communication will tend to include also social aspects. Although the cost of virtual meetings is lower than face-to-face meetings, there is a place for improvements in order to increase productivity. Virtual communication is not inadequate, but it needs to be more organized in order to be more efficient.
4. The Research Model Phrases.N
SocVsTask.P
PosVsNeg.P
QuesVsAns.P
Effectiveness Virtuallity Level Efficiency Structural Level Satisfaction
Figure 1. The Research Model
Based on the literature review and the interviews, a research model was developed as depicted in Figure 1. According to the model there are two major affecting variables, the virtuallity level and the structural level. These two affect four mediators. One of them is the number of total phrases, a measure for the discussion length. The other three can be described as social versus and task characteristics of the discussion: social versus task ratio, positive versus negative ratio and questions versus answers ratio. Finally, the measurable (dependent) variables are the outputs of the team work: efficiency, effectiveness and satisfaction. The research model is partially based on the literature review, since as far as we know, there is no previous research combining the variables virtuallity, structure and social versus task characteristics of the discussion. Therefore, in order to complete the model, we had to rely on the interviews described above, and on our own intuition. Cramton & Webber (2005) refer to the limitation of information evolves from using different types of technologies and to the increase in hostility and aggression of the communicators toward each other. We argue that distant communication is a potential source of complications due to a lack of tangible information. Therefore, we assume a negative correlation between virtuallity and both the mediators and the output variables. Powell, Piccoli & Ives (2004) consider team structure as the missing link in the research of virtual teams. In this particular research we attempt to tackle their challenge, and hope we find positive effects of structure on the discussion content (the mediators) and the team output. One of our main conclusions from the in-depth interviews, was that virtual team managers and members believe that the form of communication (i.e.: virtual or face-toface) will determine whether the discussion will be task or socially oriented. Using Bales (1950) Interaction Process Analysis (IPA) we defined four mediators, which indicate the discussion orientation. We predict that the discussion orientation is negatively influenced by the virtuallity level and positively influenced by the structural level. We also predict that the discussion orientation will impact team output. The hypotheses driven from the above model are described in the following section.
5. The research hypotheses This section briefly explains the major hypotheses emerged from the research model illustrated above. Organizing the team's work should enhance its outputs. Adopting structural characteristics is one way of teams to achieve this target. Therefore: Hypothesis no. 1 – for an intellective task, a structured team's output (with a formal Division of Labor, Hierarchy and Work Process) is superior to an unstructured team's output (without a formal Division of Labor, Hierarchy and Work Process).
The virtual working environment complicates the team work. The negative influence of virtuallity has been already extensively investigated (Cascio, 2000 ; Cramton & Webber, 1999 ; Jarvenpaa & Leidner, 1999 ; Straus & McGrath, 1994 ; Thompson & Coovert, 2003). Thus, according to previous research: Hypothesis no. 2 – for an intellective task, a virtual team's output is inferior to a face-to-face team's output (for the same task). Assuming that the structural characteristics will reduce the negative influence of virtuallity, we word an interaction hypothesis: Hypothesis no. 3 – for an intellective task, a virtual-structured team's output is superior to a virtual-unstructured team's output. The group of hypothesis no. 4 concerns with the impact of structure on the content of the team discussion: Hypothesis no. 4a – for an intellective task, structured team's discussion is longer than unstructured team's discussion. Hypothesis no. 4b – for an intellective task, structured team's discussion is task oriented, while unstructured team's discussion is social oriented. Hypothesis no. 4c – for an intellective task, the social part of the structured team's discussion is positive oriented, while the social part of the unstructured team's discussion is negative oriented. Hypothesis no. 4d – for an intellective task, the task part of the structured team's discussion is answers oriented, while the task part of the unstructured team's discussion is questions oriented. The group of hypothesis no. 5 concerns with the impact of virtuallity on the content of the team discussion: Hypothesis no. 5a – for an intellective task, virtual team's discussion is shorter than face-to-face team's discussion. Hypothesis no. 5b – for an intellective task, virtual team's discussion is task oriented, while face-to-face team's discussion is social oriented. Hypothesis no. 5c – for an intellective task, the social part of the virtual team's discussion is negative oriented, while the social part of the face-to-face team's discussion is positive oriented. Hypothesis no. 5d – for an intellective task, the task part of the virtual team's discussion is question oriented, while the social part of the face-to-face team's discussion is answer oriented. The group of hypothesis no. 6 concerns with the impact of the interaction between virtuallity and structure on the content of the team discussion, Assuming that the structural characteristics will reduce the negative influence of virtuallity: Hypothesis no. 6a – for an intellective task, virtual-structured team's discussion is longer than virtual-unstructured team's discussion. Hypothesis no. 6b – for an intellective task, virtual-structured team's discussion is task oriented, while virtual-unstructured team's discussion is socially oriented. Hypothesis no. 6c – for an intellective task, the social part of the virtual-structured team's discussion is positive oriented, while the social part of the virtual-unstructured team's discussion is negative oriented. Hypothesis no. 6d – for an intellective task, the task part of the virtual-structured team's discussion is answers oriented, while the task part of the virtual-unstructured team's discussion is questions oriented. The group of hypothesis no. 7 concerns with the impact of the content of the team discussion, namely the discussion length and the social versus and task characteristics of the discussion, on the teams output:
Hypothesis no. 7a – for an intellective task, the length of the discussion influences the team's output: a positive influence on Effectiveness, a negative influence on Efficiency and Satisfaction. Hypothesis no. 7b – for an intellective task, the Social versus Task orientation of the discussion influences the team's output: a positive influence on Satisfaction, a negative influence on Effectiveness and Efficiency. Hypothesis no. 7c – for an intellective task, the Positive versus Negative orientation of the social part of the discussion influences the team's output: a positive influence on Satisfaction, Effectiveness and Efficiency. Hypothesis no. 7d – for an intellective task, the Questions versus Answers orientation of the task part of the discussion influences the team's output: a negative influence on Satisfaction, Effectiveness and Efficiency. The next section describes the methodology used in order to test our hypotheses. The results will follow.
6. Methodology An experiment was designed, in which a team task was given to 150 undergraduate students in an academic college. The subjects, who are grouped into teams, have to share information in order to complete the task. Subjects were assigned to teams of three. Each team was given a task that takes approximately 30 minutes to complete. The research design is a 2x2 Between Subjects Factorial Design: the factors are team structure (structured vs. unstructured) and type of communication (virtual vs. face-to-face). The research design includes a total of four experimental conditions (see Table 2). Each condition was implemented on 12-13 teams, thus the experiment included 150 subjects (4 conditions * 12-13 teams * 3 subjects). Virtuallity Level Structural Level Remarks 1 0 0 Face-to-face-unstructured team 2 0 1 Face-to-face-structured team 3 1 0 virtual-unstructured team 4 1 1 virtual-structured team Table 2. Experimental Conditions
Procedure Subjects were invited in groups of three, to meetings (virtual or to non virtual) that were conducted using MSN-Messenger or face-to-face communication. correspondingly. The process of the unstructured condition of the experiment includes an intellective task. Each team member received a discrete piece of information, and only the aggregation of all the information revealed the whole "picture" and led to the correct solution. The structured condition of the experiment included also a preliminary manipulation, namely: instructions in writing to the team members asking them to nominate a chairperson who will be in charge of the assembling of the information and to appoint a spokesperson and an information coordinator for the team. In addition, a recommended work procedure was given to the team members.This preliminary manipulation was not present in the non-structured condition
Operationalization of the Dependent Variables and mediators Efficiency- the time required to complete the task. Effectiveness- the team's solution compared to the correct solution. Satisfaction- team members' reaction to the task will be measured by their understanding of communication, and satisfaction of medium, results and process. Task and Social Processes
A textual (or audio) recording was saved for each virtual (or face-to-face) meeting. Task and social processes was measured by content analysis: The analysis, from each meeting, included the number of social positive phrases compared to the number of social negative phrases; and the numbers of task question phrases compared to the number of task answer phrases, accordingly to Bales (1950) model. In order to use reliable measures, the phrase counting was done separately by two independent judges. The two judgment analysis was compared one to the other, and in a case of different decision (concerning the phrase category), a new agreed decision was taken. Four measures were calculated out of the above phrases counting: Total phrases – number of total phrases during the meeting. Social Vs Task Level – social phrases percentage related to task phrases percentage, among total phrases. Positive Vs Negative Social Level – positive phrases percentage related to negative phrases percentage, among social phrases. Questions Vs Answers Task Level – question phrases percentage related to answer phrases percentage, among task phrases. Table 3 presents a summery of the Means and SD's of all the mediating variables, above the experiment condition (Means and SD's by the independent variables are described in the following section).
M Phrases.N
SD
282.74
164.01
SocVsTask.P
24.50
9.08
PosVsNeg.P
71.71
15.74
QuesVsAns.P
28.64
5.95
Table 3. Means and SD's of the Mediator Variables (Overall N=50)
6. Findings Fifty experiments where preformed among undergraduate students in an academic college. Each one of the fifty teams included three participants, a total of 150 participants. The mean age of the participants was 25.79 years old (SD = 3.94). Among the participants, there were 89 males and 61 females. No significant differences were found among the teams in mean age (F 3 = 0.676 , p = .568). Breslow-Day test showed that there were no gender ratio differences among the experimental cells (χ 2 1 = 0.766 , p = .381). A manipulation check demonstrated that the three dimensions (hierarchy, division of labor and work process) are well manipulated in the experiment group. However, most of the team members in the control groups admitted that they had an organized work process, even if they where not given specific instructions to do so. An analysis of variance (ANOVA) was performed for assessing the effects of the independent variables (Virtuallity and Structure) on the processes and outputs variables.
Table 4 presents a summery of the Means and SD's. of all the output and mediating variables by the independent variables. Virtual team
Structured team
team
M
SD
86.53
26.25
78.47
29.40
82.66
27.52
Efficiency
41.85
12.54
24.08
8.45
33.32
13.91
Satisfaction
3.86
.53
4.25
.36
4.05
.49
177.77
79.95
450.92
193.48
308.88
199.38
SocVsTask.P
24.99
13.16
28.72
3.98
26.78
9.87
PosVsNeg.P
66.50
19.12
77.04
8.18
71.56
15.57
QuesVsAns.P
29.42
6.77
27.83
4.15
28.66
5.61
Effectiveness
69.23
28.74
84.00
24.73
76.32
27.39
Efficiency
38.62
11.55
17.58
6.21
28.52
14.12
Satisfaction
3.46
.46
4.32
.35
3.87
.59
172.38
52.59
347.83
97.49
256.60
117.22
SocVsTask.P
20.81
7.08
23.73
8.45
22.22
7.74
PosVsNeg.P
66.11
16.60
78.11
13.86
71.87
16.23
QuesVsAns.P
28.17
6.91
29.13
6.04
28.63
6.39
Effectiveness
77.88
28.37
81.23
26.72
79.49
27.36
Efficiency
40.23
11.93
20.83
7.98
30.92
14.08
Satisfaction
3.66
.52
4.28
.35
3.96
.54
175.08
66.36
399.38
158.81
282.74
164.01
SocVsTask.P
22.90
10.57
26.23
6.94
24.50
9.08
PosVsNeg.P
66.30
17.55
77.58
11.14
71.71
15.74
QuesVsAns.P
28.80
6.73
28.48
5.11
28.64
5.95
Phrases.N
Total
Phrases.N
N
13
13
26
M
SD
Total
Effectiveness
Phrases.N
Unstructured
Face-to-face team N
12
12
24
M
Table 4. Means and SD's of All the Output and Mediator Variables by the Independent Variables
SD
N
25
25
50
The ANOVA indicates, for each hypothesis respectively, that: H1. Structured teams are not more successful than unstructured teams in completing the task; structured teams are not more satisfied and do take longer time in carrying out the task (F 1,46 = 2.890 , p = .096 , η2 = 0.059). H2. Virtual teams are not less successful than face-to-face teams in completing the task; yet, virtual teams are less satisfied (F 1,46 = 25.119 , p < .001 , η2 = 0.353) and take longer time in carrying out the task (F 1,46 = 45.933 , p < .001 , η2 = 0.500). H3. Virtual-structured team members are more satisfied than virtual-unstructured team members (F 1,46 = 3.446 , p = .070 , η2 = 0.070); yet, virtual-structured teams do not take longer time in carrying out the task. A significant interaction effect was found for the independent variable structure on effectiveness among the virtual teams (U 13, 13 = 55.0 , p (1 tailed) = .044). A similar effect was not found among the face-to-face teams. H4a.H4b.H4c.H4d. Structured team's processes have no different orientation than unstructured team's processes: the discussion length is not significantly different between structured and unstructured teams; the level of social versus task is not different; the level of positive versus negative is not different; and the level of questions versus answers is not different. H5a. The total number of phrases during the virtual meetings was much lower than the total number of phrases during the face-to-face meetings (F 1,46 = 46.118 , p < .001 , η2 = 0.501). H5b. Virtual team's social versus task discussion orientation is not different than face-to-face team's social versus task discussion orientation. H5c. The social part of the virtual team's discussion is different than the social part of the face-to-face team's discussion: The positive phrases percentage related to the negative phrases percentage among social phrases of virtual teams was significantly lower (F 1,46 = 6.912 , p = .012 , η2 = 0.131). H5d. Virtual team's question versus answer discussion orientation is not different than face-to-face team's questions versus answers discussion orientation. H6a.H6b.H6c.H6d. Virtual-structured team's processes have no different orientation than virtual-unstructured team's processes: the discussion length is not significantly different between virtual-structured and virtual-unstructured teams; the level of social versus task is not different; the level of positive versus negative is not different; and the level of questions versus answers is not different. A Pearson Correlation analysis was conducted in order to measure the correlations between the (social and task) processes of the teams and the teams outputs. The analysis was done for the whole sample (N=50), above the experiment condition. A summery of the results is presented in Table 5. Phrases.N
SocVsTask.P
PosVsNeg.P
QuesVsAns.P
Effectiveness -.199
-.240
.350*
-.159
-.235
.095
-.480**
.073
.322*
-.174
.398**
-.170
Efficiency Satisfaction ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).
Table 5. Pearson Correlations between the Mediators and the Output Variables (Overall N=50)
The Pearson Correlations indicate that among all teams, there is (H7a.) a significant positive linear correlation between total amount of phrases and satisfaction (r = .322 , p = .023); (H7c.) a significant positive linear correlation between Positive Vs Negative Social Level and Effectiveness (r = +.350 , p = .013); a significant negative linear correlation between Positive Vs Negative Social Level and Efficiency (r = -.480 , p < .001); and a significant positive linear correlation between Positive Vs Negative Social Level and Satisfaction (r = +.398 , p = .004). Correlations between Social Vs Task Level and team output (H7b) were not found. Correlations between Questions Vs Task Answers Task Level and team output (H7d) were not found either. Since the focus of this study is virtual-structured teams, we also preformed a Pearson Correlations for that specific experiment condition, namely the virtual-structured teams. A summery of the results is presented in Table 6. Phrases.N
SocVsTask.P
PosVsNeg.P
QuesVsAns.P
Effectiveness .012
-.586*
.700**
-.217
.244
.456
-.339
.309
-.195
-.785**
.694**
-.321
Efficiency Satisfaction ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).
Table 6. Pearson Correlations between the Mdiators and the Output Variables for Virtual-Structured Teams Only (N=13) The Pearson Correlations indicate that among virtual-structured teams there is (H7b.) a significant negative linear correlation between Social Vs Task Level and Effectiveness (r = -.586 , p = .035); a significant negative linear correlation between Social Vs Task Level and Satisfaction (r = -.785 , p = .001); (H7c.) a significant positive linear correlation between Positive Vs Negative Social Level and Effectiveness (r = +.700 , p = .008); and a significant positive linear correlation between Positive Vs Negative Social Level and Satisfaction (r = +.694 , p = .009). Correlations between total amount of phrases and team output (H7a) were not found. Correlations between Questions Vs Task Answers Task Level and team output (H7d) were not found either.
7. Discussion Structure does not increase effectiveness or satisfaction of the teams. Though structured teams are slightly more successful than unstructured teams in completing the task and structured team members are slightly satisfied than the unstructured team members, these effects are not significant. From the virtual level perspective, the effectiveness of the virtual teams was not different from the effectiveness of the face-to-face teams, yet, the two other outcomes of the virtual teams were different: virtual team members were less satisfied than face-toface team members and less efficient. These evidences are consistent with previous studies about the negative influence of virtuallity on team work, as mentioned in the literature review. The interaction between structure and virtuallity yields a significant effect on effectiveness and satisfaction. Therefore, structure does overcome two negative impact of the virtuallity - less effectiveness and less satisfaction. Yet, it does not raise the virtual team efficiency. Concerning efficiency, it is obvious that forming a structure takes time, thus, in the short run it can not contribute to the team efficiency.
Structure has no significant effect on social and task processes. The lack of simple structure effect on the teams' output and the existence of the interaction effect on effectiveness, imply that a further attention should be given to the operational definition of structure, as a variable with three components: division of labor, hierarchy and work process. Most of the team members in the control group admitted that they had an organized work process, although they where not given specific instructions to do so. A practical conclusion is that in order to improve virtual teams' effectiveness, as opposed to traditional teams one should focus on hierarchy and division of labor rather than on work process, which is carried out anyway by the team members. This is probably due to the disadvantage of virtual teams - the transfer of information is limited. However, this may be transformed into an advantage. For example: gender, age and ethnic background of team members are recognized by traditional team members who see one another and can clearly identify their gender, age, and ethnic background. It is possible that the division of labor and hierarchy in traditional teams are based on stereotypes of team members. For example, a young white male may be appointed as a team leader instead of an older female from a minority group, which might be more suitable for the position. This lack of information in virtual teams can lead to the division of labor and hierarchy which is based on professional criteria related to the successful completion of the task. The virtuallity yields some significant effects on the content of the team discussions. First, the content analysis shows that virtual discussions are much shorter than face-to-face discussions, implying that the written word is "costly" than the spoken one. While during traditionally talking, we freely spread task and social phrases, the use of computer technologies forces us to consider each and every idea and suggestion, and to use them in a stingy manner. Second, the social part of the virtual team's discussion is negative than the social part of the face-to-face team's discussion. This evidence is also consistent with previous studies, such as Cramton & Webber (2005) insights about technology tools limitation of information, and the increase in hostility and aggression of the communicators. The correlation between the (social and task) processes of the teams and the teams output, over the entire sample, demonstrates a significant power of the Positive Vs Negative Social Level of the team discussion. Positive (versus negative) discussion orientation is translated into enhanced effectiveness, member satisfaction, and efficiency (it takes less time to carry out the task). Surprisingly, there is a positive correlation between the total number of phrases and the team members' satisfaction, suggesting that communicators' satisfaction is positively influenced by the length of the discussion. The correlation under the specific experiment condition of virtual-structured teams partially duplicates the findings shown regarding the Positive Vs Negative Social Level of the team discussion over the entire sample. Positive (versus negative) discussion orientation is translated into enhanced effectiveness and members' satisfaction. The efficiency of virtual-structured teams is not affected by the Positive Vs Negative Social Level. For the very same virtual-structured teams, the Social Vs Task Level of the discussion influences negatively the team effectiveness and its member satisfaction. This finding implies that social (versus task) discussion orientation is translated into enhanced effectiveness and members' satisfaction. The efficiency of virtual-structured teams was not affected by the Social Vs Task Level.
Limitations
Powell, Piccoli & Ives (2004) consider team structure as the missing link in the research of virtual teams. In this particular research we attempted to tackle their challenge, but with partial success. First, we would like to refer to the research method. Experiments, by definition are done in a controlled environment. Furthermore, the subjects in our experiment where college students, a group that does not necessarily reflect the participants in virtual teams in organizations. There is a wide spectrum of research methods that can be used to examine the research model, such as surveys and case studies. We believe that the examination of the proposed research model with additional research methods will assist in reinforcing it. Second, we operationally defined team structure by three dimensions: hierarchy, division of labor and work process. It is possible and recommended to define team structure through more dimensions such as levels of autonomy, formality, shared language, goals, norms and team diversity. Third, the experiments we conducted were performed by using a specific task (intellective task) with a specific form of communication (MSN messenger). Straus & McGrath (1994) argue that forms of communications that transfers more social clues will increase performance and satisfaction. It is recommended to examine the research model using additional forms of communications and tasks.
Summary The field of team structure refers to the work of traditional teams and needs to be adjusted to the 21st century and its communication media, specifically to the virtual environment. The motivation for this study was to examine the impact of e-infrastructure on the work place environment and to meet some of its challenges, hoping to contribute a new theoretical and practical solution to the wide adoption of e-infrastructure across organizations. We offered a new approach of improving virtual teams to assist organizations in the era of globalizations, an era in which many organizations need to employ virtual teams that are distance from one another. Our findings indicate, like some previous studies, that virtual teams output is inferior to face-to-face teams output: virtual team members are less satisfied than face-toface team members and less efficient. A clue for that evidence might be lying in the team's discussion orientation: the virtual team's discussion orientation is negative than the face-to-face team's discussion orientation, while Positive (versus negative) discussion orientation is translated into enhancive effectiveness, member satisfaction, and efficiency (it takes less time to carry out the task). In other word, trying to enhance the social part of the virtual discussion should increase the teams' output. The interaction between structure and virtuallity yield a significant effect on effectiveness and satisfaction. Therefore, structure does overcome two negative impact of the virtuallity - less effectiveness and less satisfaction. Yet, it does not contributes to the team efficiency. It means that we can successfully improve virtual team effectiveness and satisfaction, by swift structure, but with limited success. Only two of the three variables improved due to swift structure: virtual teams with swift structure were more successful in completing the task and the team members were more satisfied, but the efficiency (time) of team performance were not affected. An examination of the manipulation reveals that the improvement is due to the division of labor and hierarchy, and less because of work process that takes place anyway (even without our manipulation). Our explanation is that the limited information transfer that usually creates a shortcoming in virtual teams, actually worked in favor of the virtual teams in this instance by reducing stereotype images of the team members, thus "forcing" team member to take into account relevant criteria in completing the task.
Finally, we have also found that the written word is more "costly" than the spoken one. The use of computer technologies forces us to consider each and every idea and suggestion, and to use them in a miserly manner. Another interesting insight indicates that people enjoy communication: communicators' satisfaction is positively influenced by the length of the discussion. In the pervious section, we outlined future directions of virtual team research. It is likely that future studies in these directions will assist in improving virtual team effectiveness.
8. References Ahuja, M. K., Galletta, D. F. & Carley, K. M. (2003) 'Individual Centrality and Performance in Virtual R&D Groups: An Empirical Study', Management Science, 49(1), 21-38. Andres, H. P. (2002) 'A Comparison of Face-to-Face and Virtual Software Development Teams', Team Performance Management: An International Journal, 8(1/2), 39-48. Baker, G. (2002) 'The Effects of Synchronous Collaborative Technologies on Decision Making: A Study of Virtual Teams', Information Resources Management Journal, 15(4), 79-93. Bales, R. F. (1950) 'A set of factors for the analysis of small group interaction', American Sociological Review, 15, 257-263. Bouas, K. S. & Arrow, H. (1996) 'The Development of Group Identity in Computer and Face-toface Groups with Membership Change', CSCW, 4, 153-178. Boyett, J. H. & Conn, H. P. (1992). Workplace 2000: The revolution reshaping, Plume. Bradner, E., Mark, G. & Hertel, T. D. (2005) 'Team size and technology fit: participation, Awareness, and rapport in distributed teams', Professional Communication, 48(1), 68- 77. Bunderson, J. S. & Sutcliffe, K. M. (2002) 'Comparing alternative conceptualizations of functional diversity in management teams: Process and performance effects', Academy of Management Journal, 45(5), 875-894. Burke, K., Aytes, K. & Chidambaram, L. (2001) 'Media Effects on the Development of Cohesion and Process Satisfaction in Computer-Supported Workgroups', Information Technology & People, 14(2), 122-141. Burke, K. & Chidambaram, L. (1996) 'Do Mediated Contexts Differ in Information Richness? A Comparison of Collocated and Dispersed Meetings', In Proceedings of the 29th Annual Hawaii International Conference on System Sciences, Hawaii, USA, 92-101. Burke, K. & Chidambaram, L. (1995) 'Developmental Differences Between Distributed and Faceto-face Groups in Electronically Supported Meeting Environments: An Exploratory Investigation', Group Decision and Negotiation, (4), 213-223. Cascio, W. F. (2000) 'Managing a virtual workplace', The Academy of Management Executive, 14(3), 81-91. Chidambaram, L. & Tung L. L. (2005) 'Is Out of Sight, Out of Mind? An Empirical Study of Social Loafing in Technology-Supported Groups', Information Systems Research, 16(2), 149-171. Cramton, C. D. & Webber, S. S. (2005) 'Relationships among geographic dispersion, team processes, and effectiveness in software development work teams', Journal of Business Research, 58(6), 758-765. Cramton, C. D. & Webber, S. S. (1999) 'Modeling the Impact of Geographic Dispersion on Work Teams'. Working Paper, George Mason University, Washington, DC, USA. Cummings, J. N. (2004) 'Work Groups, Structural Diversity, and Knowledge Sharing in a Global Organization', Management Science, 50(3), 352-364. DeSanctis, G., Wright, M. & Jiang, L. (2001) 'Building a Global Learning Community', Communications of the ACM, 44(12), 80-82.
Graetz, K., Boyle, E., Kimble, C., Thompson, P. & Garloch, J. (1998) 'Information Sharing in Face-to-face, Teleconferencing, and Electronic Chat Groups', Small Groups Research, 29(6), 714-743. Gruenfeld, D. H., Mannix, E. A., Williams, K. Y. & Neale, M. A. (1996) 'Group composition and decision making: How member familiarity and information distribution affect process and performance', Organizational Behavior and Human Decision Processes, 67(1), 1-16. Hackman, J. R. (2002) 'Leading teams : setting the stage for great performances', Boston, Mass. : Harvard Business School Press. Hackman, J. R. (1987) 'The design of work teams', In J. W. Lorsch (ed.), Handbook of Organizational Behavior, 315-342. England Cliffs, NJ: Prentice-Hall. Hause, M., Last, M., Almstrum, V. & Woodroffe, M. (2001) 'Interaction Factors In Software Development Performance In Distributed Student Teams In Computer Science', In Proceedings of the ACM Conference, Boulder, CO, USA, 69-72. Hofner, S. D. (1996) 'Productive Behaviours of Global Business Teams', International Journal of Intercultural Relations, 20(2), 227-259. Huang, W. W., Wei, K.-K., Watson, R. T. & Tan, B. C. Y. (2003) 'Supporting Virtual TeamBuilding with a GSS: An Empirical Investigation', Decision Support Systems, 34(4), 359367. Iacono, C. S. & Weisband, S. (1997) 'Developing Trust in Virtual Teams', In Proceedings of the Thirtieth Hawaii International Conference on System Sciences, 2, 412-420. Ilgen, D. R. (1999) 'Teams embedded in organizations', American Psychologist, 54, 129-139. Jarvenpaa, S. L. & Leidner, D. E. (1999) 'Communication and trust in global virtual teams', Organization Science, 10(6), 791-817. Johnson, S. C., Suriya, C., Yoon, S. W., Berrett, J. V. & La Fleur, J. (2002) 'Team Development and Group Processes of Virtual Learning Teams', Computers & Education, 39, 379-393. Kayworth T. & Leidner, D. (2001/2002) 'Leadership Effectiveness in Global Virtual Teams', Journal of Management Information Systems, 18(3), 7-40. Kim, Y., Hiltz, S. R. & Turoff, M. (2002) 'Coordination Structures and System Restrictiveness in Distributed Group Support Systems', Group Decision and Negotiation, 11, 379-404. Lebie, L., Rhoades, J.A. & McGrath, J. E. (1996) 'Interaction Process in Computer-Mediated and Face-to-Face Groups', CSCW, 4, 127-152. Leavitt, H. J. (1996) 'The old days, hot groups, and managers' lib', Administrative Science Quarterly, 41(2), 288-301. Maier, N. R. F. (1967) 'Assets & Liabilities In Group Problem Solving', Psychology Review, 74(4), 239-249. Martins, L. L, Gilson, L. L. & Maynard, M. T. (2004) 'Virtual teams: What do we know and where do we go from here?', Journal of Management, 30(6), 805-835. Maruping, L. M. & Agarwal, R. (2004) 'Managing Team Interpersonal Processes Through Technology: A Task-Technology Fit Perspective', Journal of Applied Psychology, 89(6), 975-990. Olson, J., Olson, G. & Meader, D. (1997) 'Face-to-face Group Work Compared to Remote Group Work With and Without Video', In Video-mediated Communication, K. Finn, A. Sellen, & S. Wilbur, (Eds.), Lawrence Erlbaum Associates: Mahwah, USA, 157-172. Olson, J. & Teasley, S. (1996) 'Groupware in the Wild: Lessons Learned from a Year of Virtual Collocation', In Proceedings of the ACM Conference, Denver, CO, USA, 419-427. Potter, R. E. & Balthazard, P. A. (2002a) 'Understanding Human Interaction and Performance in the Virtual Team', Journal of Information Technology theory and Application, 4(1), 1-23.
Powell, A., Piccoli, G. & Ives, B. (2004) 'Virtual Teams: A Review of Current Literature and Directions for Future Research', Database for Advances in Information Systems, 35(1), 637. Siegel, J., Dubrovsky, V., Kiesler, S. & McGuire, T. W. (1986) 'Group Processes in ComputerMediated Communication', Organizational Behavior and Human Decision Processes, 37(2), 157-188. Straus, S. G. & McGrath, J. E. (1994) 'Does the medium matter? The interaction of task type and technology on group performance and member reactions', Journal of Applied Psychology, 79(1), 87-98. Strijbos, J. W., Martens, R. L., Jochems, W. M. G. & Broers, N. (2004) 'The Effect of Functional Roles of Group Efficiency: Using Multilevel Modeling and Content Analysis to Investigate Computer-supported Collaboration in Small Groups', Small Group Research, 35(2), 195229. Thompson, L. F. & Coovert, M. D. (2003) 'Teamwork Online: The Effects of Computer Conferencing on Perceived Confusion, Satisfaction, and Postdiscussion Accuracy', Group Dynamics: Theory, Research, and Practice, 7(2), 135-151. Tyran, K.L., Tyran, C.K. & Shepherd, M. (2003) 'Exploring Emerging Leadership in Virtual Teams', In Gibson, C. B., & Cohen, S. G. (Eds.), Virtual Teams that Work, San Francisco, CA: Jossey-Bass, 183-195. Warkentin, M., Sayeed, L. & Hightower, R. (1997) 'Virtual Teams Versus Face-to-Face Teams: An Exploratory Study of a Web-Based Conference System', Decision Sciences, 28(4), 975-996. Weisband, S. & Atwater, L. (1999) 'Evaluating Self and Others in Electronic and Face-to-Face Groups', Journal of Applied Psychology, 84(4), 632-639. Weisband, S. P., Schneider, S. K. & Connolly, T. (1995) 'Computer-Mediated Communication and Social Information: Status Salience and Status Differences', The Academy of Management Journal, 38(4), 1124-1151. Yoo, Y. & Alavi, M. (2004) 'Emergent leadership in virtual teams: What do emergent leaders do?', Information and Organization, 14(1), 27-58.
