Coefficient Beta, Item Clustering and The Homogeneity of Summated Scales

Abstract

Summated scales are widely used in management research to assess constructs of interest, such as job satisfaction, organisational culture and organisational commitment. The adequacy of summated scales can be evaluated using criteria such as reliability, validity and factorial homogeneity. We explore the utility of Revelle's (1979) coefficient beta as a criterion for judging the internal homogeneity of summated scales over and above the more typically applied coefficient alpha measure of internal consistency. The present paper applies coefficient beta, in the context of Revelle's (1978) ICLUST item clustering procedure, on responses to Warr's (1990) affective well-being scale from two distinct samples.