Electronic Thesis and Dissertation Repository


Doctor of Philosophy




Dr. Albert Carron


The general purpose of this dissertation was to develop an inventory designed to measure cohesion in children’s (ages 9-12) sport teams. To this end, three studies were conducted. In Study 1, children became active agents in the process of test construction. More specifically, children (N = 167) participated in focus groups and completed open-ended questionnaires in order to provide information on their perceptions of cohesion as well as motives for participating, continuing, and ceasing involvement on sport teams. Study 2 involved the use of the information obtained from Study 1 to develop potential items for the questionnaire. In addition, the questionnaires were distributed to child sport participants (N = 298) to determine factorial validity. Finally, the purpose of Study 3 was to establish construct validity for the Child Sport Cohesion Questionnaire (CSCQ) with a heterogeneous sample of children (N = 290). To accomplish this task, four separate tests of validity were assessed: convergent, discriminant, predictive, and factorial.

The results from Study 1 indicated that children as young as 9 years can understand the cohesion construct. They are able to identify (a) advantages relating to cohesive teams, (b) disadvantages relating to non-cohesive teams, and (c) methods for increasing cohesion within teams. Finally, in Study 1, children advanced motives for (a) joining, (b) maintaining, and (c) ceasing involvement on sport teams.

Study 2 resulted in the development of potential items for the questionnaire. The research team (N = 4) and age appropriate children (N = 8) provided content validity for the advanced items. Furthermore, factorial validity was demonstrated via confirmatory factor analysis (CFA). The resulting Child Sport Cohesion Questionnaire containing 16 items (task cohesion = 7, social cohesion = 7, negative spurious items = 2) and a 5-point Likert scale, demonstrated a strong model fit, good internal consistency values, and a moderate inter-factor correlation.

The findings from Study 3 revealed that the newly developed CSCQ possessed (a) convergent validity, (b) discriminant validity, and (c) factorial validity. In addition, partial support for predictive validity was established. Overall, the combination of the results from Study 3 provides support for the construct validity of the CSCQ.