The process of validating a recently developed instrument to assess perceived team cohesion is discussed. The Group Environment Questionnaire (GEQ), an instrument designed to measure cohesion in sport teams, has good estimates for its internal consistency and for its content and factorial validity (Carron, Widmeyer, & Brawley, 1985; Widmeyer, Brawley, & Carron, 1985). However, other aspects of its validity required examination. The present article reports three studies concerning inspection of the GEQ's concurrent (Study 1), predictive (Study 2), and construct (Study 3) validities. In Study 1 the GEQ exhibited the predicted correspondence with similar measures of cohesion and was not significantly correlated with measures of other constructs. In Study 2 the GEQ successfully discriminated team and individual sport athletes by predicting their membership to these groups on the basis of their task cohesion scores. As well, classification of athletes as new and long-standing members of individual sport teams was predicted on the basis of their social cohesion scores. Finally, in Study 3 evidence was obtained for the predicted difference in self-responsibility attributions between high and low task-cohesive athletes of team sports. Considering the results of the three studies with previous evidence of content and factorial validity, the conclusion was that the GEQ is valid. In sum, demonstrations of the GEQ's content, factorial, concurrent, predictive, and construct validity reflect the ongoing process of its construct validation.
Each investigator contributed significantly to the research for this article and considers its authorship as joint.
Requests for reprints should be sent to Lawrence R. Brawley, Department of Kinesiology, University of Waterloo, Waterloo, Ont., N2L 3G1, Canada.