Who benefits from virtual collaboration? The interplay of team member expertness and Big Five personality traits

Collaborative problem-solving (CPS) is a crucial 21st-century skill, and its virtual implementation has grown significantly due to the COVID-19 pandemic. While research exists on virtual communication processes, understanding the impact of individual attributes on virtual collaboration outcomes remains limited. This study addresses this gap by investigating the role of individual expertise and Big Five personality traits in virtual CPS team performance. Individual expertise, encompassing both the type and level of knowledge (expertise and expertness diversity), is central to CPS success. Transactive Memory Systems (TMS) theory explains how teams manage knowledge, emphasizing expertise recognition, specialization, and coordination. However, existing research often neglects other individual attributes, such as personality traits. This research proposes integrating TMS theory with the Big Five personality traits model (openness, conscientiousness, extraversion, agreeableness, and neuroticism) to examine their influence on virtual CPS performance. The study aims to understand how individual expertness, personality traits, and their diversity within teams affect performance gains through TMS mechanisms, leading to a richer understanding of successful virtual CPS.

Transactive Memory Systems (TMS) theory focuses on how teams distribute and manage knowledge, highlighting the roles of expertise recognition, specialization, and coordination. Research shows that expertise diversity (differences in expertise across domains) positively impacts team performance. However, expertness diversity (differences in expertise levels within a domain) is less studied. Existing TMS research often overlooks other individual attributes, notably personality traits, which are known to influence team performance and collaboration. The Big Five personality traits model provides a framework to examine these traits (openness, conscientiousness, extraversion, agreeableness, and neuroticism). However, the interaction between personality traits and TMS development in virtual CPS settings is under-researched. This study aims to bridge this gap by integrating TMS and Big Five models to explore virtual CPS team dynamics.

This experimental study used 377 ad hoc dyadic teams (754 participants) recruited through Amazon Mechanical Turk. Participants completed a general science knowledge test (to assess expertness), the Ten Item Personality Inventory (TIPI) to measure Big Five personality traits, and a demographic survey. Teams then completed a virtual simulation-based collaborative task involving a general science quiz answered through online synchronous text chat. Individual and team performance gains were measured by comparing scores before and after collaboration. Individual expertness was categorized into high (H) and low (L) based on the median test score. Expertness diversity was categorized as high (LH, one H and one L member) and low (LL, both L members; HH, both H members). For personality traits, each member was classified as H or L based on the median score for each trait. Diversity for each trait was similarly categorized as high and low. ANCOVAs analyzed the influence of expertness and personality diversity on team and individual performance gain, controlling for initial performance. Post-hoc Tukey's tests examined subgroup differences.

At the team level, high expertness diversity significantly improved performance (LH teams > LL and HH teams). Teams with low agreeableness diversity (both members low in agreeableness) showed the greatest performance gains. At the individual level, low-expertness members collaborating with high-expertness partners (LwH) demonstrated the most significant performance improvement. Agreeableness showed a significant effect in the full model for individual performance, but post-hoc tests showed no significant differences among the subgroups. The reliability of the general science knowledge test was high (Cronbach's α = 0.89).

The findings support the integration of TMS and Big Five models in understanding virtual CPS. High expertness diversity enhances team performance, contradicting some previous research and highlighting the potential benefits of reciprocal learning between high- and low-expertness individuals in virtual settings. The positive impact of low agreeableness diversity suggests that direct communication and a focus on task completion can be advantageous in virtual environments where social cues are reduced. The significant individual-level effect of expertness underscores the importance of appropriate team composition for maximizing virtual collaboration benefits. These findings extend TMS theory by demonstrating the positive influence of expertness diversity in virtual CPS and the significant contributions of the Big Five personality traits, particularly agreeableness, in shaping team dynamics and outcomes.

This study demonstrates the value of integrating TMS theory and the Big Five personality traits model to understand virtual team dynamics. High expertness diversity and low agreeableness diversity are associated with greater performance gains in virtual collaborative problem-solving. Low-expertness individuals benefit significantly from collaboration with high-expertness teammates. Future research should investigate longer-term collaborations, larger team sizes, and incorporate communication content analysis. The study’s practical implications highlight the importance of strategic team formation to maximize virtual collaboration benefits.

The study's limitations include the use of ad hoc dyadic teams, a short collaboration duration, text-only communication, and a limited task scope. The generalizability to real-world settings might be limited. Future research should address these limitations by investigating more natural collaborations with longer durations, larger teams, and richer communication modalities. Further, the study focuses primarily on expertness and personality traits; other individual attributes could be considered in future research. Also, analysis did not include the text-chat content, which could provide additional insights.