2012 45th Hawaii International Conference on System Sciences

A Member - Virtual Team Fit Theory: Group Effects on Performance

Yi (Jenny) Zhang , Ph.D. Ofir Turel, Ph.D. California State University, California State University, Fullerton, CA Fullerton, CA [email protected] [email protected]

Abstract importance of input (e.g., team design) variables as Virtual teams have become common in many predictors of team performance. For example, it has business settings. Thus, it is important to examine been shown that team size [8], and the cultural factors that affect their performance. Although much diversity of teams influence (albeit, sometimes research has been done to this end, the influence of indirectly) virtual team outcomes [9]. virtual team personality composition on team One relatively unexplored design attribute of performance has rarely been studied. In this study, we virtual teams is (i.e., configuration) seek to bridge this gap by theoretically developing and in terms of the personalities of team members. empirically testing a model that captures these effects. Understanding the impacts of this team attribute on Data collected from 61 virtual teams suggest that the team performance would allow the development of elevation (median) level of team conscientiousness optimal team staffing strategies, e.g., by allocating improves team performance, while higher within-team optimal combinations of individuals to work in virtual diversity (standard deviation) in conscientiousness teams. Although several studies highlight the potential lowers performance. Furthermore, the elevation level importance of this team design consideration for of team conscientiousness moderates the negative performance [3, 10-12], there has been little empirical effect of within-team conscientiousness diversity on research in this area. Our study therefore intends to team performance. Implications for research and start bridging this gap. We do so by extending person- practice are discussed. traditional team fit theories [13, 14] to virtual settings. The latter type of teams can be challenging to manage due to communication and team process difficulties, 1. Introduction which are less salient in face-to-face teams [2-5]. The advance our understanding of virtual teams, we The nature of tasks in modern organizations has investigate the influence of their personality become more complex, dynamic, and global. In configurations on team performance. We focus on a response, virtual teams that heavily rely on information single trait of team members, namely their technologies for team interactions have become conscientiousness, because this trait is often considered prevalent in a broad range of business settings and for highly influential in team contexts [13]. It can strongly a plethora of tasks [1]. These teams have the following impact key teamwork behaviors (e.g., leadership, characteristics: they include several individuals who sharing, social loafing) [15]. Focusing on virtual team may not be collocated and who rely on technology for composition in terms of conscientiousness, we much of their communication; and in these teams examine whether it influences team performance, and individuals work on interdependent tasks, and share the mechanisms through which various aspects of the responsibility for outcomes [2]. composition, such as elevation (central tendency) and Given the increased importance and relevance of diversity (variation) influence virtual team virtual teams, they have attracted the attention of performance. Information Systems (IS) scholars. One focal outcome To this end, we develop hypotheses based on of interest in this line of research has been team traditional person-environment fit theories. The performance [2-5] . It is crucial to understand what hypotheses are then tested and validated with data drives the performance of virtual teams in order to collected from 187 team members nested in 61 virtual increase their vitality and usefulness. To this end, past teams in an academic setting. Implications for virtual research have mostly looked at how social, cognitive team research and practice are then discussed. and emotional process variables, such as trust [6]; and Ultimately, this study advances a member- virtual team how task process factors, such as conflict [1] or fit theory, which is lacking from current works on coordination [7], influence team performance. virtual teams. Furthermore, past studies have recognized the

978-0-7695-4525-7/12 $26.00 © 2012 IEEE 829 DOI 10.1109/HICSS.2012.64

2. Theoretical Background conscientiousness of its members, and the effect of this composition on virtual team performance. This focus was guided by theory. First, individuals who are 2.1. Personality Configuration conscientious are achievement oriented, competent, The five-factor, also known as the “big Five” orderly and deliberate [16]. In virtual teams it is model (FFM) is a widely accepted personality difficult to track other members’ activities, intentions, framework. The FFM includes the following five plans and current status because face-to-face contact elemental traits: Extraversion (being sociable, assertive, and immediate feedback are lacking [25]. Thus, being talkative, and active), Agreeableness (being courteous, self-motivated, task oriented and staying on schedule flexible, trusting, good-natured, cooperative and could be a facilitating condition for functioning well in tolerant), Conscientiousness (being careful, thorough, virtual teams. responsible, and organized), Emotional Stability Second, being competent and responsible also (opposite of being anxious, depressed, angry, worried, helps gaining interpersonal trust and consequently, and secure), and Openness to Experience (being reducing potential disagreements [30]. Not surprisingly, imaginative, curious, original, broad-minded, and several meta-analysis have demonstrated that intelligence) [16-18]. conscientiousness is the strongest personality trait This framework and others have been used to associated with performance [31], and that it can study the effectiveness of various personality positively affect a range of team processes and configurations in face-to-face teams. The foci have outcomes[32, 33]. been on the effects of elevation (i.e., a composite team- Third, the effect of personality traits on level attribute, such as average level of a trait) and performance often depends on the nature of the task. diversity (i.e., dispersion of team-level attribute, such For example, extraversion appears to influence as the standard deviation of within-team levels of a outcomes when tasks involve considerable interaction trait), on team performance [19]. While these effects with customers [34]. Conscientiousness, however, has have been studied in traditional team settings [20-22], been found to be related to performance regardless of we argue that the trait composition of virtual teams task types [31]. That is, the effects of deserves special attention. conscientiousness on team outcomes generalize well, On the one hand, virtual teams and traditional and are context independent [35]. teams face different communication challenges. For example, in virtual teams, the lack of face-to-face 2.3. Person-Environment Fit contact leads to reduced ability to convey non-verbal The life cycle model of virtual teams categorizes cues and to lack of immediate response in virtual team related variables into inputs, socio- communications [23, 24]. As such, virtual teams find it emotional processes, task processes, and outputs [4]. In more difficult to become cohesive and to perform well this study, we examine how an input variable, namely [25], often present heightened levels of social loafing [8], and frequently struggle to build trust and team configuration (or composition), affects an output variable, namely team performance. We further build relationships among team members [6, 26, 27], which on person-environment fit theories to explain the are crucial for team performance and other outcomes mechanisms through which this effect takes place [14]. [28]. On the other hand, virtual environments present The person-environment (P-E) fit paradigm [36, less social pressure, so people feel more free to express their ideas [29]. Thus, personalities that excel in face- 37] posits that the congruence between team member to-face settings may fail in virtual settings, and vice characteristics and the team environment can influence individual behavior in team contexts. Ultimately, the fit versa. For instance, a creative but introvert person may between these elements is a driver of team outcome not contribute ideas in face-to-face meetings, but may variables, such as satisfaction, effectiveness, and feel more comfortable to do so in online settings. As performance [14]. This proposition has been supported such, findings from traditional team research may not generalize well to virtual teams. in traditional team settings [38, 39]. Studies also focused on team member conscientiousness, both at the individual [31] and team [15] levels. At the team level, 2.2. Why conscientiousness? which is the focus of this study, the findings suggest From a practical standpoint, taking into account all that composite team-level conscientiousness is a key traits may be too cumbersome and overly complicated. predictor of team performance [15]. But, findings Thus, a good initial step, and a proof of concept, would regarding conscientiousness diversity effects have been be to focus on a single but influential trait. Specifically, mixed. we focus only on team composition in terms of the Two different types of fit are normally discussed in the P-E fit literature. One is based on the notion of

830 complementary fit which occurs when the needs of the indirectly through encouraging other members to be environment are fulfilled by the unique strengths of more responsive and goal oriented [46]. Past research individuals [36]. Complementary fit is therefore the has also shown a positive relationship between team match between the demand and supply between the members’ achievement (an aspect of work environment and individual. The other type of fit conscientiousness) and team performance and is supplementary fit. Supplementary fit occurs when an efficiency [47, 48]. individual and the working environment possess These views are further encapsulated in person- similar or matching characteristics. The focus on role fit theories [49], according to which a team will be studies embracing the supplementary P-E fit more effective when there is a match between the perspective has been mostly on value congruence personality characteristics and the team member roles. between employees and organizations [40]. In the Conscientious members are adaptable to any role current study, because the focus is on one trait requirement of the teams [19, 44], and as such, often fit (conscientiousness), the supplementary fit perspective well to any team in any role. Indeed, it has been shown is taken. that the overall level of conscientiousness in a team Supplementary fit is manifested through the predicts performance across task types [31]. Hence: within-team variation in conscientiousness (diversity). H1: Within-virtual team elevation of conscientiousness Since conscientiousness reflects one’s value position, will positively affect team performance. reducing conscientiousness variation in a team will increase supplementary fit through goal congruence. As demonstrated above, the within-team level of Members of such teams will have similar views conscientiousness should positively influence regarding what they want to achieve and how much performance. This team attribute, however, portrays effort to put forward. This common background will only one dimension of team configuration. The other reduce conflict caused by disagreement regarding facet we examine in this study is the within-team effort, commitment, and goals. diversity in conscientiousness. Building on the supplementary fit perspective [36], we argue that 3. Research Model within team diversity in conscientiousness , i.e., having individuals with different levels of conscientiousness in the same team, is detrimental to team performance. Studies of traditional teams have shown that a The supplementary fit perspective posits that higher elevation of conscientiousness in a team leads to people are more comfortable and productive when they better performance [41, 42]. Members of such high are similar to others [50]. In team settings, elevation teams are easier to trust, because their supplementary fit exists when team members share behaviors are more predictable and are consistent with similar goals and values [13]. Supplementary fit works the higher standards of practice. These team members through several interrelated mechanisms to drive better expect others to take the task seriously, put effort into team performance. First, team members will use the it, and do their part in a timely manner. shared goals and values for enhancing interpersonal High conscientiousness is especially important in trust [51]. Trust is an important process factor that virtual teams, in which it is harder to establish trust and helps virtual teams to perform better [52]. When teams avoid destructive conflict [11]. Highly conscientious have high supplementary fit all members have similar team members are a partial remedy to these problems. goal orientation and commitment, and they all work Conscientiousness can be characterized by two diligently towards the team objective. This caters to dimensions: dependability and achievement-orientation characteristic-based trust production [53]. Furthermore, [43]. Being dependable makes conscientious sharing similar goals and values with teammates also individuals self-disciplined, organized, and very reduces the unpredictability in communications which responsive. These attributes can be crucial to the is common in virtual teams [11]. Thus, it can lead to success of virtual teams, as in such settings, individual higher trust, lower levels of interpersonal conflict, and team members have less control over what others do. ultimately to high performance [54-56]. In contrast, if they suspect that others do not prioritize Second, supplementary fit ensures that all group team-related tasks and put forth their best effort, they members share norms regarding the effort they need to may reduce their own effort, which will lead to a put into the task. This is important in virtual teams, as downward spiral of social loafing [8]. team members exert no or little control on the actions Furthermore, highly conscientious members are of others, and this state of “out of sight” promotes achievement-oriented. They will exert effort and social loafing behaviors [8]. A supplementary fit based perseverance toward goal accomplishment [44, 45]. shared perspective on effort and timeline norms Through their persistence and hard-work, not only that ensures that people do not engage in social loafing, and they contribute to performance directly, but also that they do not perceive others to engage in social

831 loafing [57]. The latter is very important, because assume that, at least to some extent, the less based on equity theory [58], it guarantees that conscientious team members will adapt, and present individuals feel that their efforts are weighted and actual attitudes and behaviors that are better than these rewarded properly [59, 60], which can prevent and predicted by their basic levels of conscientiousness. Of reduce social loafing behaviors [61]. course, there are also cases where people will not cave In contrast, a team with members who hold to peer-pressure and accordingly, will not change their different levels of conscientiousness should attitudes. For example, individuals who feel that they underperform. First, in such teams individuals will are dispensable in the team will exert less effort [72]. have different opinions on how much they want to In such cases, because highly conscientious members accomplish and how much effort to put into the task. are the majority, they may be more willing to absorb High conscientious members aspirate to achieve high the work of the social loafers to ensure the team performance since they are achievement-oriented. Low outcome. Overall, when a majority of highly conscientious members, on the other hand, do not conscientious members exists, the negative impact of value high performance as much as their high the diversity (dissimilarity) in conscientiousness will conscientiousness peers. This by itself is a basis for be mitigated. within-team conflict, which undermines team On the other hand, when a low conscientiousness performance [62]. Second, when the less conscientious majority exists, the prevalent expectation would be for team members exert less effort, the others will likely low performance, and these members will likely follow, as it would seem to them unfair that they put remain quiet and inactive [60]. Because the highly more effort, but receive the same reward. [63]. This conscientious members are the minority, they would equity-based downward social-loafing spiral of effort try to prevent to play the “sucker” role, and reduce is harmful to team performance [64]. Thus: their effort to reflect the group effort [73, 74]. Thus, in such cases we may observe a stronger than before H2: Within-virtual team diversity of conscientiousness negative effect of conscientiousness diversity on team will negatively affect team performance. performance. While the effects of team conscientiousness The abovementioned points may be exemplified elevation (composite team conscientiousness) have using these hypothetical teams. One three-member been successfully demonstrated in face-to-face team team has the following levels of conscientiousness: settings [19, 35, 65], findings regarding the effect of 1,7,7 and the other has these levels: 7,1,1. The conscientiousness diversity on team performance have variation in these teams is the same (standard deviation been mixed. Some studies have shown a significant = 3.46). Nevertheless, as argued above, different effect [e.g. 35], and some demonstrate no significant behaviors may be expected in such teams. In the first, relationships [e.g. 65]. We thus postulate that there the majority is highly conscientious, in which case they may be an interaction effects that can explain such can influence the less conscientious member to follow discrepancies. their example. In the second, the majority is less In line with advances in person-environment fit conscientious, in which case the downward spiral of theory [66], it is plausible that the composite level of social loafing is more likely to occur. Hence: within-team conscientiousness moderates the influence H3: The elevation of within-virtual team of within-team diversity in conscientiousness, such that conscientiousness will moderate the effect of within- the negative impact of within-team variation will be virtual team variability in conscientiousness on team mitigated. This moderation effect stems from attitude performance, such that when the elevation is high the modification processes that prevail in teams with high negative impact of variability will be mitigated. within-team differences, in which there is naturally a majority and minority subgroups. According to social 4. Research Methodology identity theory [67] and social influence theories [68, 69] individuals may change their attitudes and consequent behaviors because of informational or 4.1. Sample and Procedure normative influences [70]. Research participants were 290 undergraduate When a majority of highly conscientious business students in an IS course, who were randomly individuals exists, they will aggressively pursue their assigned to mostly three-person virtual teams (90 out goal, and are likely to attempt to persuade the less of 95 groups) and a few four-person virtual teams (5 conscientious individuals to put more effort into the out of 95 groups). They were asked to use a private task. Based on social identity theory [67], less discussion board to discuss a real-world case-study, conscientious individuals will change their cognition over a period of four weeks, and then write a report. and behavior in part to maintain positive social or The case study described the privacy and security shared group identity [71]. Thus, it is reasonable to practices of Apple’s iPhone. The discussion board

832 allowed teammates to discuss the case, exchange ideas deviation, or mean Euclidean distance [77], out of and files (e.g., MS Word documents with draft which, for simplicity reasons, we selected to use the analyses), and develop the final submission. first one. Note that the common practice is to use either Participants were explicitly asked to work only via variance or standard deviation for capturing diversity their assigned virtual collaboration spaces, and not to as separation [35]. Since the distribution of variance is meet face-to-face or use alternative collaboration tools known to be potentially skewed because of its higher (e.g., personal email accounts). Following assignment order, standard deviation is used in this study as the conclusion, students were asked to complete a paper- measure of within-team trait variability or dissimilarity. based survey which captured demographic information and their levels of conscientiousness. Voluntary 5. Data Analysis & Results participation was encouraged with a small course credit. A total of 246 responses were obtained (85% First, the reliability of the conscientiousness response rate), but 59 pertained to groups for which we measures that were taken at the individual level was had only one or two responses. Because this study assessed. A Cronbach’s alpha of 0.89 indicated that the focuses also on diversity measures, only responses scale was reliable. Because all our hypotheses pertain from groups with at least 3 responses were retained. to teams (i.e., are at the team-level), the team-level Two responses per group may not be enough because constructs were operationalized using median and the variance may be easily deflated or inflated, and standard deviation measures applied to single-response groups do not allow variance conscientiousness scores from each team. Second, assessment. Ultimately, a sample of 187 responses that descriptive statistics of and correlations among the pertained to groups for which we had at least 3 team-level constructs were assessed (see Table 1). responses was retained. These responses were nested in 57 three-person teams and 4 four-person teams (a total Table 1. Descriptive statistics and correlations of 61 teams, net response rate of 65%).

4.2. Operationalization of Constructs The survey included a well-established conscientiousness measures [75] which used a one to seven Likert scale anchored with “strongly disagree” (1) and “strongly agree” (7). Team performance was measured as the team project grade as assigned by the course instructors, independently from this study. * p<0.1, ** p<0.05 Measures of conscientiousness were taken at the individual level (i.e., from each team member), and Linear regression was used to test all the then following the composite-construct typology hypotheses. First, two simple linear regression models specified by Chan [76], were aggregated to the team- were estimated to test the separate main effects of level. Traditionally researchers tend to use the mean elevation and variability of conscientiousness on team value as an aggregate measure of elevation, e.g., performance (see Table 2). Barrick et l. [34]. However, since most of the groups in Table 2. Simple linear regression results our study consist only of three members, using mean for these purposes would be very sensitive to the Coefficients Model extreme values. As a result, median is chosen as a Statistics Estimates central location measure because of its insensibility to Model extreme values. The within-team median is further β p-Val R2 advantageous as the elevation measure because median, b unlike mean, is not involved in formulating variance. 1. Elevation .260** .067 .068** Within-team variability was operationalized as a Performance separation construct [77] , i.e., one that is extreme b when there is a bimodal distribution, and minimal 2. Variation -.236* .042 .056* when the distribution is narrow and uni-modal. While Performance measures of diversity as variety or disparity exist as * p<0.1, ** p<0.05 well, separation is most appropriate in our case because it best captures disagreements and opposing attitudes The significance of the coefficient of elevation of (and beliefs regarding task importance, in our case). conscientiousness in model 1 lends support to H1. The Such measures are best operationalized with standard negative impact of variability in conscientiousness on

833 team performance (H2) is marginally supported (p = As can be seen from the above equation, the 0.067) by model 2. That is, virtual teams with members negative impact of within-team variability in who are highly conscientious perform better than conscientiousness is alleviated by higher levels of others, and so are virtual teams with members who elevation of conscientiousness (above 5.52 in this possess similar degrees of conscientiousness. particular case). For example, when the elevation is To test the interaction effect in H3, hierarchical average (elevation variable=0 in the equation), the multiple regression with mean-centered predictor effect of the variation on performance is -2.158; and variables [78] was conducted. Centering reduces when the elevation level is 4 (elevation variable=-1.52 Variance Inflation Factor (VIF) to acceptable levels of in the equation), the effect of the variation on multicollinearity. A general rule is that the VIF should performance is -14.48. not exceed 10 [79]. The main effects were tested in To further illustrate the moderation effect of the step-1. Then, the cross-product term was added to test elevation level of team conscientiousness, we plotted the interaction hypotheses in step-2. See results in the relationship between within-team variability in Table 3. conscientiousness and performance at the mean as well as at other levels of elevation (plus 3 to minus 3 Table 3: Hierarchical regression standard deviations). Figure 1 also provides the one- tailed significance levels of the lines. As can be seen, the relationship between within-team variation and performance depends on the within-team elevation level of conscientiousness. At higher elevation level, the negative effect of within-team variation on performance is alleviated. At lower elevation levels of Note. N=61 teams. Unstandardized regression conscientiousness, the negative effect of the within- coefficients are reported (with standard errors in team variation in conscientiousness on performance is parentheses) amplified. The lower the elevation level, the steeper * p<0.1, ** p<0.05 the slope becomes. These results further confirm the hypothesis about the moderation effect of the elevation The interaction effect hypothesis (H3) is supported level of conscientiousness, and provide more nuanced by the significant product term in the step-2 model. It insights. is also noticeable that R-square has increased dramatically by adding the interaction effect (0.172 compared to 0.087), which is significant at the .05 level. With the interaction effect, 17% of the variance in team performance can be explained by just team level conscientiousness. The VIF for all the centered predictor variables are very low. Thus multicollinearity is unlikely to be a problem. This result is considerably higher than what others managed to explain in other personality- team performance studies. For example Barrick et al [35] managed to explain only 6.8% of the variation in team performance based on the mean level of conscientiousness, and 10.9% of variance in team performance based on the within-team variance in conscientiousness. One possible reason for their lower R2 values is the omission of the interaction effect. In order to explicate the moderation effect, the full regression equation for step-2 model in Table 3 is given below:

Team Performance = 85.553+1.286Elevation -2.158Elevation + 7.366Elevation*Variation =85.553+1.286Elevation - (2.158 - 7.366Elevation)*Variation 1

1 The coefficients are unstandardized coefficients β instead of the standardized coefficients β in Table 3.

834 achievement orientation and dependableness, are also important for the success of virtual teams. Third, the diversity of conscientiousness within a team was negatively related to team performance. This finding is consistent with the supplementary fit perspective of P-E fit where minimized diversity will lead to better performance. Thus, this theory and its extensions are likely to be relevant also in virtual settings. Lastly, we found that the negative effect of conscientiousness diversity on performance is moderated by the elevation of conscientiousness. The negative effect was strong and significant when the elevation of conscientiousness was low, and it was mitigated when the elevation of conscientiousness was high. It implies that people may behave differently based on differences in their team environments. When the environment has many conscientious members, less conscientious individuals may change their attitudes and become more cooperative and put in more effort to achieve the general group goal. On the other hand, when the environment includes mostly low conscientiousness members, individuals may lower their expectations and efforts to match others in the group. This is a possible explanation to inconsistencies in past research regarding this effect [e.g., 65]. Yet, it Figure 1: Virtual team conscientiousness elevation deserves more research. moderates the effect of conscientiousness variation on performance 6.2. Implications for Practice

6. Discussion Traditional team staffing strategies focus mostly on individual personality traits. Our results indicate that the personality composition of the team matters. This study’s results show that not only did the Hence, managers should assign high conscientious elevation and variation of conscientiousness directly individuals to virtual teams to increase team influence team performance, but their interaction also conscientiousness elevation. They should also assign had a strong impact on it. Specifically, the elevation individuals who are similar in terms of level moderates the relationship between within-team conscientiousness to a team. However, if the majority conscientiousness diversity and team performance. of team members are high in conscientiousness, adding This helps explain the inconsistent findings in previous less conscientious individuals would not significantly studies regarding the effects of personality diversity on hurt the team’s effectiveness. On the other hand, team performance. dysfunctional teams with low levels of

conscientiousness may be “cured” by adding to them a 6.1. Implications for Research large enough number of highly conscientious Our findings provide several insights for virtual individuals. team research. First, conscientiousness elevation explained 17% of the variance in team performance. It 6.3. Limitations and Future Research confirms the importance of personality-based staffing Several limitations should be acknowledged. First, strategies [11], and suggests that future research should the results may apply to the specific task type we include this team attribute as an important predictor of studied. Other task types should be examined in the performance. future. Second, this study used student teams, which Second, the elevation of conscientiousness of a may not generalize well to other virtual teams. Future team was found to be positively related to team research may replicate our findings in organizational performance. This finding is consistent with findings in settings. Third, we focused on one personality trait. traditional teams. It confirms that the attributes Depending on the task and the amount of interaction associated with conscientiousness, such as required, other traits may also be predictive of team

835 outcomes, and should be considered in future research. integration," Journal of Applied , vol. Fourth, this study only focused on the direct impact of 89, pp. 822-834, 2004. personality traits on performance which did not include [14] J. R. Edwards, D. M. Cable, I. O. Williamson, L. S. any behavior and process variables that could serve as Lambert, and A. J. Shipp, "The phenomenology of fit: Linking the person and environment to the mediators between personality traits and performance. subjective experience of person-environment fit," Journal of Applied Psychology, vol. 91, pp. 802- 7. Reference 827, 2006. [15] M. A. G. Peeters, H. Van Tuijl, C. G. Rutte, and I. [1] R. L. Wakefield, D. E. Leidner, and G. Garrison, Reymen, "Personality and team performance: A "A Model of Conflict, Leadership, and meta-analysis," European Journal of Personality, Performance in Virtual Teams," Information vol. 20, pp. 377-396, Aug 2006. Systems Research, vol. 19, pp. 434-455, Dec 2008. [16] P. T. Costa and R. R. McCrae, Revised NEO [2] J. Webster and S. Staples, "Comparing virtual Personality Inventory(NEO-PIR): Professional teams to traditional teams: An identification of new manual. Odessa, FL: Psychological Assessment research opportunities," in Research in Personal Resources, 1992. and Human Resources Management. vol. 25, J. J. [17] J. M. Digman, "Personality structure: Emergence Martocchio, Ed. Boston: Elsevier, 2006, pp. 181- of the five-factor model," Annual Review of 214. Psychology, vol. 41, pp. 417-440, 1990. [3] L. L. Martins, L. L. Gilson, and M. T. Maynard, [18] L. R. Goldberg, "The development of markers of "Virtual teams: What do we know and where do we the Big-Five factor structure," Psychology go from here?," Journal of Management, vol. 30, Assessment, vol. 4, pp. 26-42, 1992. pp. 805-835, 2004. [19] G. A. Neuman and J. Wright, "Team effectiveness: [4] A. Powell, C. Piccoli, and B. Ives, "Virtual Teams: Beyond skills and cognitive ability," Journal of A Review of Current Literature and Directions for Applied Psychology, vol. 84, pp. 376-389, Jun Future Research," The DATA BASE for Advances 1999. in Information Systems, vol. 35, pp. 6-36, 2004. [20] B. D. Edwards, E. A. Day, W. Arthur, and S. T. [5] D. S. Staples and J. Webster, "Exploring traditional Bell, "Relationships among team ability and virtual team members' "best practices" - A composition, team mental models, and team social cognitive theory perspective," Small Group performance," Journal of Applied Psychology, vol. Research, vol. 38, pp. 60-97, Feb 2007. 91, pp. 727-736, May 2006. [6] D. L. Paul and R. R. McDaniel, "A field study of [21] P. Mallette and R. L. Hogler, "Board composition, the effect of interpersonal trust on virtual stock ownership and the exemption of directors collaborative relationship performance," MIS from liability," Journal of Management, vol. 21, pp. Quarterly, vol. 28, pp. 183-227, Jun 2004. 861-878, 1995. [7] P. Kanawattanachai and Y. Yoo, "The impact of [22] K. H. Price, D. A. Harrison, and J. H. Gavin, knowledge coordination on virtual team "Withholding inputs in team contexts: Member performance over time," MIS Quarterly, vol. 31, pp. composition, interaction processes, evaluation 783-808, Dec 2007. structure, and social loafing," vol. 91, pp. 1375- [8] L. Chidambaram and L. L. Tung, "Is out of sight, 1384, Nov 2006. out of mind? An empirical study of social loafing [23] L. Sproull and S. Kiesler, "Reducing social context in technology-supported groups," Information cues: Electronic mail in organizational Systems Research, vol. 16, pp. 149-168, 2005. communication," Management Science, vol. 32, pp. [9] D. S. Staples and L. Zhao, "The effects of cultural 1492-1512, 1986. diversity in virtual teams versus face-to-face [24] L. Sproull and S. Kiesler, Connections: New Ways teams," Group Decision and Negotiation, vol. 15, of Working in the Networked Organization. pp. 389-406, 2006. Cambridge, MA: MIT Press, 1991. [10] C. Gillam and C. Oppenheim, "Review article: [25] E. de Pillis and K. Furumo, "Counting the cost of Reviewing the impact of virtual teams in the virtual teams," Communications of the ACM, vol. information age," Journal of Information Science, 50, pp. 93-95, Dec 2007. vol. 32, pp. 160-175, 2006. [26] C. Ridings, D. Gefen, and B. Arinze, "Some [11] A. Powell, G. Piccoli, and B. Ives, "Virtual Teams: Antecedents and Effects of Trust in Virtual A Review of Current Literature and Directions for Communities," Journal of Strategic Information Future Research," Database for Advances in Systems, vol. 11, pp. 271-295, 2002. Information Systems, vol. 35, pp. 6-36, 2004. [27] A. Abdul-Rahman and S. Hailes, "Supporting trust [12] S. Z. Schiller and M. Mandviwalla, "Virtual team in virtual communities," in System Sciences, 2000. research - An analysis of theory use and a Proceedings of the 33rd Annual Hawaii framework for theory appropriation," Small Group International Conference on, 2000, p. 9 pp. vol.1. Research, vol. 38, pp. 12-59, Feb 2007. [28] C. Lin, C. Standing, and Y. C. Liu, "A model to [13] D. M. Cable and J. R. Edwards, "Complementary develop effective virtual teams," Decision Support and supplementary fit: A theoretical and empirical Systems, vol. 45, pp. 1031-1045, Nov 2008.

836 [29] S. P. Weisband, "Group discussion and 1st Research in personnel and human resource advocacy effects in computer-mediated and face- management. vol. 13, K. M. Rowland and G. Ferris, to-face decision-making groups," Organizational Eds. Greenwich, CT: JAI Press, 1995, pp. 153-200. Behavior and Human Decision Processes, vol. 53, [44] J. A. LePine, "Team adaptation and postchange pp. 352-380, Dec 1992. performance: Effects of team composition in terms [30] M. Becerra and A. K. Gupta, "Perceived of members' cognitive ability and personality," trustworthiness within the organization: The Journal of Applied Psychology, vol. 88, pp. 27-39, moderating impact of communication frequency on Feb 2003. trustor and trustee effects," Organization Science, [45] E. Molleman, A. Nauta, and K. A. Jehn, "Person- vol. 14, pp. 32-44, Jan-Feb 2003. job fit applied to teamwork - A multilevel [31] M. R. Barrick, M. K. Mount, and T. A. Judge, approach," Small Group Research, vol. 35, pp. "Personality and performance at the beginning of 515-539, Oct 2004. the new millennium: What do we know and where [46] C. Greene, "Cohesion and productivity in work do we go next?," International Journal of Selection groups," Small Group Behavior, vol. 20, pp. 70-86, and Assessment, vol. 9, pp. 9-30, Mar-Jun 2001. 1989. [32] L. M. Hough, "The 'Big Five' personality variables [47] F. W. Schneider and J. G. Delaney, "Effect of - construct confusion: Description versus individual achievement motivation on group prediction," Human Performance, vol. 5, pp. 139- problem-solving efficiency," Journal of Social 155, 1992. Psychology, vol. 86, pp. 291-298, 1972. [33] M. K. Mount, M. R. Barrick, and G. L. Stewart, [48] A. Zander and J. Forward, "Position in group, "Five-factor model of personality and performance achievement-motivation, and group aspiration," in jobs involving interpersonal interactions," Journal of Personality and Social Psychology, vol. Human Performance, vol. 11, 1998. 8, pp. 282-288, 1968. [34] M. Barrick and M. Mount, "The big five [49] M. E. Shaw and B. Harkey, "Some effects of personality dimensions and job performance: A congruency of member characteristics and group meta-analysis," Personnel Psychology, vol. 44, pp. structure upon group behavior," Journal of 1-26, 1991. Personality and Social Psychology, vol. 34, pp. [35] M. R. Barrick, G. L. Stewart, M. J. Neubert, and M. 412-418, 1976. K. Mount, "Relating member ability and [50] C. J. Resick, B. B. Baltes, and C. W. Shantz, personality to work-team processes and team "Person-organization fit and work-related attitudes effectiveness," Journal of Applied Psychology, vol. and decisions: Examining interactive effects with 83, pp. 377-391, 1998. job fit and conscientiousness," Journal of Applied [36] P. M. Muchinsky and C. J. Monahan, "What is Psychology, vol. 92, pp. 1446-1455, Sep 2007. person-environment congruence? Supplementary [51] A. S. Tsui and C. A. O'Reilly, "Beyond simple versus complementary fit," Journal of Vocational demographic effects: The importance of relational Behavior, vol. 31, pp. 268-277, 1987. demography in superior-subordinate dyads," [37] A. L. Kristof, "Person-organization fit: An Academy of Management Journal, vol. 32, pp. 402- integrative review of its conceptualizations, 423, 1989. measurement, and implications," Personnel [52] B. A. Aubert and B. L. Kelsey, "Further Psychology, vol. 49, pp. 1-49, 1996. understanding of trust and performance in virtual [38] R. B. Cattell, "Concepts and methods in the teams," 2003, pp. 575-618. measurement of group syntality," Psychological [53] L. G. Zucker, "Production of Trust: Institutional Review, vol. 55, pp. 48-63, 1948. Sources of Economic Structure, 1840-1920," in [39] W. Haythorn, "The influence of individual Research in . vol. 8, B. M. members on the characteristics of small groups," Staw and L. L. Cummings, Eds. Greenwich, CN: Journal of Abnormal and Social Psychology, vol. JAI Press, 1986, pp. 53-111. 48, pp. 276-284, 1953. [54] D. Antonioni and H. Park, "The effects of [40] A. L. Kristof, "Person-organization fit: An personality similarity on peer ratings of contextual integrative review of its conceptualizations, work behaviors," Personnel Psychology, vol. 54, measurement, and implications," Personnel pp. 331-360, Sum 2001. Psychology, vol. 49, pp. 1-49, 1996. [55] C. K. W. De Dreu and L. R. Weingart, "Task [41] M. A. G. Peeters, C. G. Rutte, H. van Tuijl, and I. versus relationship conflict, team performance, and Reymen, "The Big Five personality traits and team member satisfaction: A meta-analysis," individual satisfaction with the team," Small Group Journal of Applied Psychology, vol. 88, pp. 741- Research, vol. 37, pp. 187-211, Apr 2006. 749, 2003. [42] M. A. G. Peeters, C. G. Rutte, H. van Tuijl, and I. [56] E. M. Engle and R. G. Lord, "Implicit theories, Reymen, "Designing in teams - Does personality self-schemas, and leader-member exchange," matter?," Small Group Research, vol. 39, pp. 438- Academy of Management Journal, vol. 40, pp. 988- 467, Aug 2008. 1010, 1997. [43] M. Mount and M. Barrick, "The Big Five [57] J. M. George, "Extrinsic and intrinsic origins of personality dimensions: Implications for research perceived social loafing in organizations," and practice in human resources management," in

837 Academy of Management Journal, vol. 35, pp. 191- [68] J. Fulk, "Social Construction of Communication 202, 1992. Technology," Academy of Management Journal, [58] J. S. Adams, "Inequity in social exchange," in vol. 36, pp. 921-950, Oct 1993. Advances in experimental social psychology. vol. 2, [69] J. Fulk, J. Schmitz, and C. W. Steinfield, "A social L. Berkowitz, Ed. New York, NY, USA: Academic influence model of technology use," in Press, 1965, pp. 267-299. Organizations and Communication Technology, J. [59] B. Latané, K. D. Williams, and S. Harkins, "Many Fulk and C. W. Steinfield Eds. Newbury Park, CA, hands make light the work: The causes and USA: Sage, 1990, pp. 117-140. consequences of social loafing," Journal of [70] W. Wood, "Attitude Change: Persuasion and Social Personality and Social Psychology, vol. 37, pp. Influence," Annual Review of Psychology, vol. 51, 822-832, 1979. pp. 539-570, 2000. [60] J. A. Shepperd and K. M. Taylor, "Social loafing [71] M. A. Hogg and D. J. Terry, "Social identity and and expectancy-value theory," Personality and self-categorization processes in organizational Social Psychology Bulletin, vol. 25, pp. 1147-1158, contexts," Academy of Management Review, vol. Sep 1999. 25, pp. 121-140, 2000. [61] J. M. Jackson and S. G. Harkins, "Equity in effort: [72] S. Harkins and R. Petty, "Effects of task difficulty An explanation of the social loafing effect," and task uniqueness on social loafing," Journal of Journal of Personality and Social Psychology, vol. Personality and Social Psychology, vol. 43, pp. 49, pp. 1199-1206, 1985. 1214-1229, 1982. [62] D. V. Day and A. G. Bedeian, "Personality [73] M. E. Schnake, "Equity in effort: The "sucker similarity and work-related outcomes among effect" in co-acting groups," Journal of African-American Nursing personnel: A test of the Management, vol. 17, pp. 41-55, 1991. supplementary model of person-environment [74] K. Williams and S. J. Karau, "Social loafing and congruence," Journal of Vocational Behavior, vol. social compensation: The effect of expectations of 46, pp. 55-70, 1995. co-worker performance," Journal of Personality [63] N. L. Kerr, "Motivation losses in small groups: A and Social Psychology, vol. 61, pp. 570-581, 1991. social dilemma analysis," Journal of Personality [75] J. C. Mowen, The 3M model of motivation and and Social Psychology, vol. 45, pp. 819-828, 1983. personality: Theory and empirical applications to [64] S. J. Karau and K. Williams, D., "Social loafing: A consumer behavior. Boston, MA: Kluwer, 2000. Meta-Analytic Review and Theoretical [76] D. Chan, "Functional relations among constructs in Integration," Journal of Personality and Social the same content domain at different levels of Psychology, vol. 65, pp. 681-706, 1993. analysis: A typology of composition models," [65] G. A. Neuman, S. H. Wagner, and N. D. Journal of Applied Psychology, vol. 83, pp. 234- Christiansen, "The relationship between work-team 246, Apr 1998. personality composition and the job performance of [77] D. A. Harrison and K. J. Klein, "What's the teams," Group & Organization Management, vol. difference? Diversity constructs as separation, 24, pp. 28-45, Mar 1999. variety, or disparity in organizations," Academy of [66] S. E. Humphrey, J. R. Hollenbeck, C. J. Meyer, Management Review, vol. 32, pp. 1199-1228, 2007. and D. R. Ilgen, "Trait configurations in self- [78] L. S. Aiken and S. G. West, Multiple regression: managed teams: A conceptual examination of the Testing and interpreting interactions. Newbury use of seeding for maximizing and minimizing trait Park, CA: Sage, 1991. variance in teams," Journal of Applied Psychology, [79] D. A. Belsley, E. Kuh, and R. E. Welsch, vol. 92, pp. 885-892, 2007. Regression Diagnostics: Identifying influential [67] H. Tajfel and J. Turner, "The social identity theory data and sources of collinearity. NY, NY: John of intergroup behavior," in Psychology of Wiley, 1980. Intergroup Relations, S. Worchel and W. Austin, Eds. Chicago, IL: Nelson, 1986.

838