Abstract
Feather (1979) has pointed out the need for a measure of values applied to the work situation. This article outlines the procedures involved in the construction of such a scale. Further, it summarises the results of some of the major statistical analyses undertaken in the development of the Work Aspect Preference Scale (WAPS). An initial version of the WAPS containing 178 items was analysed by a technique recently suggested by Berven and Hubert (1977) of using factor analysis and cluster analysis as complementary multivariate techniques. Principal components analysis with a varimax rotation was repeatedly applied to samples of items to approximate simple structure. Cluster analysis confirmed the 12 factor solution. Factors were designated Security, Self Development, Altruism, Life Style, Physical Activity, Detachment, Independence, Prestige, Management, Co-Workers, Creativity and Money. Item-total correlational analysis revealed quite homogeneous subscales while a MANOVA indicated that all 12 factor scores from the highest loading 3 items significantly discriminated occupational choice categories. A second version of the WAPS, comprised of 57 items, provided very strong confirmation of the previous results with the addition of an extra hypothesised dimension Surroundings which had been inadequately represented in the earlier version. Again cluster analysis confirmed the factor structure. Item-total correlations for each factor confirmed the homogeneity of all 13 subscales and a MANOVA indicated that all three-item subscales but Prestige, significantly discriminated students' occupational choices. A third version of the WAPS was developed with the addition of an extra item per subscale in order to investigate whether four items provide more valid and reliable assessment than three. Reasons for delaying the investigation of the reliability of the subscales and the most appropriate estimate of reliability are also discussed.

This publication has 14 references indexed in Scilit: