Membership stability and performance: a study of baseball clubs and laboratory groups

Abstract

This research aims to analyse the association across membership stability, group boundary and group performance. Using Professional Chinese Baseball League's data (Seasons 2003–2007), Study One reveals that the membership stability of baseball clubs was correlated with their annual rank and win-probability, and that stable teams outperformed unstable teams. Membership stability also predicted win-probability in four out of five seasons. Using undergraduate students, Study Two reveals that members from unstable groups (i.e. groups with higher membership turnover) had higher intention of leaving their groups and more departures. Members from stable groups had better group performance, group cohesion, and group dynamics, which all resulted in members displaying more trust in their colleagues and an increased willingness to cooperate with them. Implications of the findings include that stable membership is a crucial factor in the promotion of group dynamics and overall performance, and that the nature of a group's boundary affects its membership turnover. Policies should also be devised to manage membership stability, as maintaining an appropriate turnover rate is essential both to individuals and to groups, such as employees and their organizations.

Keywords:

• boundary
• group
• membership
• performance
• stability

Notes

1. Using more complex laboratory experiments, this hypothesis will be further examined in Study Two. Group performance will be measured by its investment profits. Group dynamics will be scrutinized by a series of indicators, including: levels of cohesion, trust between group members, cooperation between group members, intention of leaving group and frequency of actual departure. Full details are discussed in Study Two.

2. Criticism often levied at the behavioural, cognitive and biological approaches due to their use of the laboratory experiment in their research. Experiments by their very nature often get human participants behaving in a non-normal (or non-natural) situation, which may lead to distorted data, and thus weak and non-convincing conclusions. Laboratory experiments are thus accused of lacking in ecological validity, i.e., the research data, analysis and implication do not reflect a real life situation. To ensure the ecological validity of the research, Study One therefore adopted real baseball clubs (as research subjects) and collected data in a natural manner. That is, performance of baseball clubs and players were both directly observable, measurable and cannot be manipulated by the researchers. All the data gathered were genuine and analysed in an objective manner, which provides ample support to the credibility of data mining and, ultimately, better ecological validity of research findings.

3. In this particular analysis, as we aimed to explore the variances of longitudinal TRN/TRL of baseball clubs across five seasons (2003–2007), we regarded TRN/TRL as dependent variables and therefore adopted Repeated-Measure ANOVA in the analysis. We are aware that other researchers may prefer GLM (General Linear Model) for the same analysis.