The following post is a review/critique of :
Barrick, M.R., Stewart, G.L., Neubert, M.J. & Mount, M.K. (1998).
Relating member ability and personality to work-team processes and team effectiveness.
Journal of Applied Psychology, 83(3), 377-391.
Barrick et. al (1998) investigated relationships between team level variables of abilities, group personality traits and social cohesion to team level viability and performance. The study offers significant advances in the understanding of teams in both theory and empirics allowing for a comprehensive assessment of how individual factors interplay and could be grouped into a team-level analysis.
There are number of impressive aspects to this paper. Intra-group processes of highly complex team processes of dynamics, interactions, and communications are not easily achieved by mere quantitative approaches, yet in this study different measures for the different constructs measured by individual level scales provide a more meaningful insight into what goes on in a team, backed by theoretical logic as to what type of information those methods reflect and how important those are in assessing the team. The study is quite rigorous in empirical methodology, settings and analysis, having being conducted in field real-life settings from several organizations, combining both employee and supervisor ratings on a very wide array of scales and factors. The analysis being conducted using the different approached to meet the research questions and hypotheses is clear and convincing, making for a very strong general appeal. I found it quite difficult to suggest major issues for improvement.
With that said, there are a few smaller issues to consider in the paper. In general, because the paper covers a wide array of topics at the same time, this results in some parts of the paper that are not sufficiently elaborated sometimes giving a sense of parsimony in that some of the hypotheses are not very clear in regards to the theory driving them. For example, it is not entirely clear why social cohesiveness is expected to mediate personal traits in hypothesis seven. The authors state what they expect would happen but the authors do not share their theoretical logic with the readers as to why that should take place.
It would have also made sense to better differentiate between theory and empirics. I personally prefer not to have empirical conceptualization of a concept in a hypothesis (stating the method – variance, means etc.). Instead, explaining the theoretical meanings of each method could be outlined first with the hypotheses building on top of that with mere reference to the theory and meaning rather than the method. This would also allow for higher clarity in terms the theoretical build-up of the arguments.
In terms of the three methods suggested, several hypotheses have been made independently for two of the methods in the same direction. While this makes sense for minimum and means, having hypotheses in terms of variance and means or variance and minimum-maximum seems at times almost contradicting (something that the authors acknowledge). For example, 4a and 4b suggest that both higher extraversion means and higher variance in extraversion both contribute to higher ratings of capability to continue working together, which may be two different directions of the role extraversion of individuals plays in the team. The results further suggest that this is the case, as 4a receives full support while 4b does not. Perhaps an interaction of the two, rather than two independent hypotheses might be in order for those combinations of team level analysis.
Relating to some of the literature we read on in terms of the conceptualization of fit, the recently popular approach is to assess how an individual fits within the group, person-group fit, comparing an individual level factor with a group level score. Given the methods described it would be interesting to try and understand how that reflects on the concept of fit. One of the methods, variance, seems to incorporate the sense of group fit to some extent comparing each score to the mean, but to my limited understanding there is a long debate on that conceptualization and how it should be used and interpreted (Edwards, 1994 ; Kristof, 1996 ; Kristof, 2005).
In terms of fit and some of the arguments made, I believe it’s also important to take into consideration that people with different traits might have different attitudes towards fit. For example, it’s been argued that extraverted people prefer to have extraverted people working with them, and so variance within the team interplays with one’s preferences towards the construction of this variance, further affecting personal and then group level outcomes. According the input-output model described in the introduction it would also make sense that a few organization level factors might reflect on how those individual and team level variables (with the interaction between the two) might have different meaning and outcomes in different settings for organizations with different goals, as for example the manager’s characteristics (as representing the organization) or the type of work required might show a preference towards a certain type of individuals and teams. Thus, arguments for the generalizability of these hypotheses and the findings should be done with extreme care. I personally see the contribution of this paper more in terms of the discussion that needs to take place on teams and the tools on how to substantiate arguments in that discussion.
In summary, a very interesting study, rich in both rigor and relevance.
