Response format and factor structure in mood adjective check lists

Abstract.— Different response formats designed for mood measurement were analyzed and compared. The interest was focused upon the effects of ( a ) symmetry versus asymmetry and ( b ) inclusion versus exclusion of a "cannot decide"-category on intercorrelation between variables and thus, factor structure. It was found that the proportions of positive and negative responses were equivalent in the symmetric format but also in the asymmetric if the "cannot decide"-category was included. Generally it was found that the asymmetric format with the "cannot decide"-category resulted in a drop of both positive and negative correlations. Data emanating from the different formats were factor analyzed and it was found that the symmetric format caused marked bipolar factors which absorbed more of common variance than those of the other formats. It was established that the symmetric format best reflected an external criterion concerning rated degree of dissimilarity between variables.