Beyond Cronbach's Alpha: Considering Confirmatory Factor Analysis and Segmentation

Cronbach's alpha is a potentially useful statistic for scale development that can be misleading when used improperly. As a result, it is important for human dimensions researchers to use the statistic cautiously. The purpose of this study is to explore the importance of segmentation and confirmatory factor analysis in scale development. Data were obtained from mail surveys on wolves and wolf management given to Grand Teton National Park visitors and Greater Yellowstone Area residents. Cronbach's alpha was calculated on the survey questions regarding emotions about wolves for park visitors and local residents merged and each segment individually. Confirmatory factor analysis was also conducted on each segment. Beyond demonstrating the importance of confirmatroy factor analysis in scale development, results suggest that segmentation is an important consideration. Indices commonly used by researchers to measure social science concepts such as emotions about wolves cannot be assumed to fit all segments.