تنوع برداشت ها: اثر متقابل تنوع منابع و فناوری ارتباطی بر عملکرد گروه خلاق

Our research is based on arguments that three different diversity sources in groups – agreeableness, openness, and ethnicity – might simultaneously possess separation properties that result in social categorization and variety properties that provide non-redundant and value-adding information resources. To help understand how these diversity sources interact with the additive and reductive features of communication technology to impact group creativity, we designed two studies involving computer mediation, nominal group technique, and face-to-face (control) communication. Our findings suggest that agreeableness, openness, and ethic diversity possess both negative separation and positive variety properties. Whereas the separation properties of all three diversity sources, as well as the variety properties of openness diversity, are evident in newly-formed groups, the variety properties of agreeableness and ethnic diversity are only manifest in mature groups. Finally, the additive and reductive features of communication technology interact with all three diversity sources to impact creative group performance in different ways.

In recent years, several interesting developments in the field of group diversity research have enriched the literature. Among these are investigations of the complex nature of diversity and its relationship to group processes and outcomes over time (e.g., Harrison et al., 2002, Watson et al., 1993 and Zellmer-Bruhn et al., 2008), research that describes communication technology interventions to facilitate diverse group performance (e.g., Carte and Chidambaram, 2004 and Staples and Zhao, 2006), and studies that examine how different sources of diversity impact group outcomes, theory, and research design (e.g., Harrison and Klein, 2007, Harrison et al., 1998, McGrath et al., 1995 and Van Knippenberg et al., 2004). Much of this progress has been motivated by the confusing results found in the group diversity literature; Milliken and Martins (1996) appropriately described diversity in groups as a “double-edged sword” when referring to these paradoxical findings. The paradox refers to the potential for diverse groups to solve complex problems requiring increased creativity, multiple perspectives, and variegated expertise (Westphal & Milton, 2000), which is offset by the fact that diverse groups often experience frustration and tension (Milliken & Martins, 1996), conflict while working on complex tasks (Jehn, Northcraft, & Neale, 1999), and low levels of satisfaction with their group experience (De Dreu & Weingart, 2003). Further inhibiting diversity’s potential are ineffective communication norms, such as failure of members to actively share their own ideas, to attend to others’ different perspectives, and to encourage others’ participation during group discussions (Bhappu et al., 1997 and Maznevski, 1994). In considering different diversity sources, those easily accessible upon initial group contact (e.g., ethnicity, age, and gender) are particularly salient markers of social identity. As a result, such surface-level diversity sources (Harrison et al., 2002) often give rise to “pernicious” social categorizations (Brickson, 2000) that subvert the creative potential of diverse groups. However, the effect of surface-level diversity sources on group effectiveness decreases over time as deep-level diversity sources (e.g., functional expertise, life experience, and personality) become more influential (Harrison et al., 2002 and Zellmer-Bruhn et al., 2008). Furthermore, if variety among group members (Harrison & Klein, 2007) on a given diversity source has task relevance (Van Knippenberg et al., 2004), then that diversity source may provide a basis for improved group effectiveness and performance. Just as task and time considerations are relevant to furthering our understanding of diversity, so too is communication technology. Different communication technologies vary to the extent that they possess reductive or additive capabilities (Carte & Chidambaram, 2004). Reductive features essentially strip out elements of the communication environment (e.g., visual and vocal cues about ethnicity, age, and gender) that often give rise to social categorizations whereas additive features (e.g., the ability for individuals to share ideas in parallel) enhance group performance. Given that diversity in groups is a double-edged sword, the potential of communication technologies to reduce social categorization while enhancing group decision-making is of particular interest to some diversity scholars (DeSanctis & Monge, 1999). This study was motivated by consideration of these and related findings in the literature on diversity in groups. We were particularly intrigued by the possibility that the additive and reductive features of communication technology might positively impact the effect of different diversity sources on group outcomes depending on the propensity of a given diversity source to elicit social categorizations and its relevance to the task at hand. Therefore, we conducted two studies to further explore this possibility. In Study 1, we ran a field experiment among student groups who had been working together for a semester. They performed a creative decision-making task using their choice of either computer-mediated (CMC) or face-to-face (FTF) communication. In Study 2, we further explored these findings using a larger sample, random assignment of individuals to newly-formed groups based on personality diversity, random assignment to communication-technology conditions, a third experimental condition of the nominal group technique (NGT) (Van de Ven & Delbecq, 1971), and a more complex creative, decision-making task.

In this section, we now integrate the reported results of both Study 1 and Study 2 while highlighting how our findings contribute to the literature that we have previously cited, followed by the limitations of the two studies, as well as directions for future research. As we will discuss at length, our study contributes to the literature by: (a) establishing perceptual awareness of personality diversity in groups, thus presenting the possibility of separation properties of personality diversity, (b) testing perceptual awareness of ethnicity diversity directly, and (c) finding interactions of diversity and communication technology, which supports both separation and variety properties of the diversity sources in our study, as well as establishing the particular value that communication technology holds for diverse groups. Our two different studies also allow us to conjecture, however tentatively, about dynamic relationships of diversity, communication technology, and creative performance in both newly-formed and mature groups. When it comes to personality diversity sources, we found that group members can detect both agreeableness and openness diversity. Perceived agreeableness and openness diversity were significantly and positively related to actual agreeableness and openness diversity, respectively, in Study 1. The actual detection of personality occurred prior to Study 1 because the groups in question had worked together for almost 3 months before Study 1 commenced and participants were asked about the perceived personality of their group members. Furthermore, the results of both Study 1 and Study 2 indicate that personality diversity sources have a meaningful influence on creative group performance. Agreeableness diversity appears to have separation properties because it had a significant and negative effect on creativity in Study 2’s in newly-formed FTF groups. This finding is consistent with previous research (Barrick et al., 1998) that found agreeableness diversity to correlate significantly and negatively with various group process-related variables. But the lack of a significant effect for agreeableness diversity on creativity in Study 1’s mature FTF groups may indicate that the negative separation properties of agreeableness diversity were actually suppressed by its positive variety properties. Zellmer-Bruhn et al.’s (2008) finding that informational diversity only manifests during later stages of group development provides some support for this plausible explanation. Interestingly, we found that the additive (process structuring) features of NGT in Study 2 were beneficial for groups diverse in agreeableness but that the additive (parallel processing) and reductive (no visual and verbal cues) features of CMC in Study 1 were actually detrimental for them. By formally separating idea generation from idea evaluation, NGT in Study 2’s newly-formed groups may have diminished the separation properties of agreeableness diversity by helping members low in agreeableness be less judgmental and those high in agreeableness share more controversial ideas, thereby enhancing their group creativity. But in Study 1’s mature groups, CMC may have augmented the separation properties of agreeableness diversity by making individuals low on agreeableness that much more critical of their group members’ ideas; parallel processing gave them unbridled voice while the lack of visual and verbal cues deindividuated their group members and made the social impact of their comments less evident. This type of CMC interaction may have alienated group members high on agreeableness and caused them to participate less during group discussions, thereby reducing the number of creative ideas generated in these groups. But what’s perplexing is why we did not find a similar significant and negative effect for agreeableness diversity on creativity in Study 2’s newly-formed, CMC groups. It may be that the disinhibiting and deindividuating environment of CMC is only consequential for groups diverse on agreeableness that have matured to where members are familiar with each other and comfortable expressing criticism. We did not find any significant main effects for openness diversity on creativity in Study 1 or Study 2. This finding is consistent with the general lack of support in the literature when it comes to linking openness diversity in groups to performance outcomes (e.g., Barrick et al., 1998, Bell, 2007 and Neuman et al., 1999). That being said, this finding may actually indicate that, in both newly-formed and mature groups, the negative separation properties of openness diversity were suppressed by its positive variety properties. This explanation is even more plausible when considering that the additive (parallel processing) and reductive (anonymity) features of CMC were beneficial for Study 2’s newly-formed groups that were diverse in openness. In CMC, anonymity may have reduced the separation properties of openness diversity by preventing individuals from categorizing their group members whereas parallel processing may have increased the variety properties of openness diversity by facilitating a healthy balance of both divergent and convergent thinking. But again, what’s puzzling is why we did not find a similar significant and positive effect for openness diversity on creativity in Study 1’s mature, CMC groups. It may be that, because openness diversity has such strong separation and variety properties, the additive and reductive features of CMC are not able to reduce social categorization or enhance information processing in mature groups that have already developed subgroups and/or ineffective communication norms as a result of their openness diversity. We also found that the additive (process structuring) features of NGT were detrimental for Study 2’s newly-formed groups that were diverse in openness. One way to interpret this finding is that the process structuring of NGT may have minimized premature closure in groups low (but not diverse) on openness while enabling groups high on openness to reach closure during group brainstorming. Similar to agreeableness diversity, we found that ethnic diversity had a significant and negative effect on creativity in Study 2’s newly-formed FTF groups but not in Study 1’s mature FTF groups. This finding suggests that the negative separation properties of ethnic diversity are stronger when FTF groups first form but that they are suppressed by the positive variety properties of ethnic diversity at later stages of group development, which is consistent with findings from previous research (e.g., Harrison et al., 2002 and Zellmer-Bruhn et al., 2008). We also found that groups diverse in ethnicity benefited from the additive (process structuring) features of NGT, as well as the additive (parallel processing) and reductive (no visual and verbal cues in Study 1 and anonymity in Study 2) features of CMC; ethnic diversity had a positive and significant effect on creative group performance in both these technology conditions. Process structuring in NGT appears to have distilled the task-relevant, variety properties of ethnic diversity by formally separating idea generation from idea evaluation. But the lack of visual and verbal cues plus anonymity in CMC may have suppressed ethnic diversity’s separation properties by diminishing social categorization and ingroup bias whereas parallel processing may have highlighted its variety properties by eliminating conversational dominance and giving voice to otherwise silent group members. Bhappu et al. (1997) made similar conclusions about the effects of CMC and gender diversity in groups. Interestingly, eight of the nine interaction terms across our studies were in the opposite direction of the main effect term. Thus, a parsimonious second, not necessarily competing explanation, for our findings is that communication technology serves not just as a performance enhancer for diverse groups as we proposed, but as a performance equalizer that generally mitigates group process losses attributable to either diversity or homogeneity. While we focused on diversity in the current studies, homogenous groups may at times spend a great deal of their energy on the social aspects of group development to the ultimate detriment of task performance. It may be that appropriate communication technologies generally shift a group’s attention towards the task. If so, a shift in attention that either reduces social cues or imposes task structure may in some contexts provide benefits when group process losses associated with homogeneity (or diversity) are present. In summary, the combined findings of our two studies suggest that group members perceive agreeableness and openness diversity, in addition to ethnic diversity. Agreeableness, openness, and ethic diversity all appear to have negative separation and positive variety properties when it comes to creative group performance. Whereas the separation properties of all three diversity sources, as well as the variety properties of openness diversity, are evident in newly-formed groups, the variety properties of agreeableness and ethnic diversity are only manifest in mature groups. Finally, the additive and reductive features of communication technology interact with all three diversity sources to impact creative group performance in different ways. A key contribution of this research is that we were able to empirically distinguish between the separation and variety properties of different diversity sources. We were also able to isolate the reductive and additive features of communication technology. In doing so, we have shown that research designs capable of varying one or more contextual factors can be very useful in teasing out and better understanding otherwise elusive effects of diversity in groups, thereby facilitating theory development. Finally, we believe that because managers today often have little control over group composition, shifting their strategic focus to things they can control, such as the additive and reductive features of communication technology, should help them better harvest the value in diversity. Key limitations of this research are those typically associated with experimental designs. Field research would allow us to study groups working together using different communication technology in an intensive nature across a range of tasks. This would allow us to understand temporal and special dynamics beyond the polar opposites considered in our current research. Future research, in addition to focusing on the field, needs to strategically consider and assess other diversity sources that vary in properties such as those identified by Harrison and Klein (2007). We believe that what we earlier described as a “sub-atomic” view of diversity is essential to unlocking diversity’s “double-edged sword” (Milliken & Martins, 1996) and helping to untangle the confusing and paradoxical findings in the diversity literature. Specifically, we believe that previous scholars have underestimated the separation properties of personality diversity. Ideally, research should systematically vary diversity sources and properties with contingencies that help to move the field forward more efficiently and effectively.