Compared the results obtained by two traditional designs on the Career Maturity Inventory-Attitude Scale with the effects of an alternative design: constructing a composite longitudinal gradient. The two traditional methodologies yielded different results from each other for males, but not for females. (Author)

Tested models of interests by examining the significance of disconfirmed ordinal predictions. Examined data regarding Holland's hexagonal model and Roe's circular ordering. Tested adequacy of the hierarchical model and compared significance of the disconfirmed predictions of the hierachical model to that of the hexagonal-circular model. (Author)