Within the realm of sport management, team identification has been examined as single dimensional construct (Wann & Branscombe, 1993). However, research in social psychology has examined group identity as a multi-dimensional concept. To accurately measure and to more fully understand the implications of team identification, the construct should be studied as a multi-dimensional construct. This study is a first attempt to examine team identity as a multi-dimensional construct. The TEAM*ID scale was developed based on the literature review of Ashmore, Deaux, and McLaughlin-Volpe (2004), who proposed nine constructs of group identity: 1) Self categorization, 2) Public evaluation, 3) Private evaluation, 4) Importance, 5) Attachment – Interconnection of self with group, 6) Attachment – Sense of interdependence, 7) Social embeddedness, 8) Behavioral involvement, and 9) Content and meaning.
The purpose of this study was to develop an instrument to measure team identity as a multi-dimensional construct. Data were collected from the fan base of a large SE college football team (N=311); confirmatory factor analysis was performed on a split sample (Sample 1=155, Sample 2=156) to examine the nine constructs of group identity. Six factors (Public evaluation, Private evaluation, Interconnection of self, Sense of interdependence, Behavioral involvement, and Awareness and knowledge) emerged from the analysis and showed a good fit of the model. A review of the scale by a panel of experts provided initial evidence of content validity; Comparing the TEAM*ID scale with a portion of the Collective Self-Esteem Scale (Luhtanen & Crocker, 1992) and a revised version of the Psychological Commitment to Team (Mahony, Madrigal, & Howard, 2000) scale provided initial evidence of nomological validity. Several factors in the model did not, however, demonstrate discriminant validity. The two attachment factors lacked discrimination, suggesting the need for further development. There were also a lack of discrimination between private evaluation, behavioral involvement and the factor awareness and knowledge of the content and meaning. The concerns with discriminant validity may be sample related, however, future refinement of the model is desirable.