Purpose -The purpose of this study is to test the relationship between team composition and team performance by applying the construct of Ten Haaf, Bikker and Adriaanse. Design/methodology/approach -Data were collected from a sample of 39 teams out of a population of 234 undergraduate MBA students. Four performance indices of each team were tested against the Ten Haaf et al. scores by using Spearman's rank test. Findings -A statistical relation between team composition according to Ten Haaf et al. and team performance was not found. It is argued that Ten Haaf et al. and other authors do not use constructs according to Belbin's theory of balancing teams. Research limitations/implications -There is need for a construct, which is isomorphic with Belbin's notions of team balancing. The relation between this construct and performance should be tested. Practical implications -The current team balancing constructs in literature are of limited use. As the various constructs give differing assessments of balance these constructs should not be used. Implications are suggested for more prescriptive algorithms for designing balanced teams. Originality/value -The team composition algorithm of Ten Haaf et al. has not been tested up to now with regard to team performance. The study amplifies on the ambiguity of the concept of team balance.

No abstract