A clarification of Holland's construct of differentiation: The importance of score elevation

In J. L. Holland's (1985, Making Vocational Choices (2nd ed.), Englewood Cliffs, NJ: Prentice-Hall) theory of vocational choice, the construct of differentiation refers to the degree to which an individual's interests are clearly defined and is operationalized in the present study as the numerical difference between an individual's highest and lowest scores on the General Occupational Theme scales (GOT) of the Strong-Campbell Interest Inventory (SCII) (D. P. Campbell and J. C. Hansen, Manual for SVIB-SCII (3rd ed.), Palo Alto, CA: Stanford Univ. Press). Techniques for measuring differentiation do not account for the elevation of an interest profile; thus, subjects with strong interests (all high scores) are treated as equivalent to subjects with weak interests (all low scores). The present study examined the custom of treating all undifferentiated subjects as if they were equal, by dividing subjects into High-Score Undifferentiated (HSU) and Low-Score Undifferentiated (LSU) groups, based on the elevation of the highest GOT score. Compared to the LSU subjects, HSU subjects (a) were more likely to have SCII profiles which were internally consistent, (b) had a higher mean cumulative grade point average, (c) had a higher mean Academic Comfort score, and (d) were more likely to persist in college. These results suggest that accounting for score elevation does reveal differences between HSU and LSU subjects and that future studies should not treat undifferentiated subjects as a homogeneous group.