Taxonomy of Vocational Interests of Academic Major Areas

Abstract

This study appraised the degree to which student profiles on the Jackson Vocational Interest Survey (JVIS) for different academic majors could be clustered in a meaningful way. From an initial sample of 10,134 students, a matrix of mean scores for 131 academic majors on each of 34 JVIS basic interest scales was computed. This matrix was subjected to a singular value decomposition with subsequent orthogonal and oblique rotations of 17 reference axes. The 17 clusters so defined reflected distinct sets of academic major fields, with separate clusters for majors in engineering, computer science, performing arts, communicative arts, human services, and others. Male and female groups entered into the definition of every cluster. Based on the salient representations of academic majors on these reference axes, modal cluster profiles were computed and decomposed into five orthogonal higher-order dimensions. The implications of these findings for the psychology of occupational choice, career development, and vocational interest measurement are outlined.This study was designed to appraise the degree to which the vocational interest profiles of students in different academic major fields in a large, diverse university can be classified into cogent clusters sharing a hierarchical organization.The traditional approach to measuring vocational interests has been to identify a particular educational or occupational group, administer a vocational interest test to the group, contrast their mean item scores with those of a like-gender general group, identify items answered differentially, and develop a scoring key to assess the similarity of a respondent's item responses to the responses of the standardization group. Notwithstanding the past successes of such an approach, difficulties have become evident recently. The use of a large number of standardization groups becomes unwieldy, the multivariate psychological bases for the similarity may not be clear, and problems are posed by (a) the necessity of producing separately keyed scales for males and for females, and (b) correlations among related occupations. We therefore wish to evaluate the degree to which a clustering approach to vocational interest measurement, using basic interest scales rather than individual items, might overcome these difficulties.

The structure of vocational interests.

“…Roe (1956) proposed eight clusters, whereas Holland (1973) preferred six. Other, more recent applications often use more scales or clusters (e.g., 10 in the Activity Experience Inventory [Ewens, 1981], 18 in the Career Decision Making System [Harrington & O'Shea, 1982], and 34 in the Jackson Vocational Interest Survey [Jackson et al, 1984]). The compatibility of the different classifications suggests that the differences in the number of clusters reflect no more than different levels of refinement.…”

“…Finally, on the third level, 207 occupational scales, representing 106 occupations, are presented. Jackson et al (1984) used a hierarchical classification approach in a study of vocational interests associated with academic majors. They found clusters that were organized hierarchically into more parsimonious, broader clusters.…”

Vocational Interests: A Look at the Past 70 Years and a Glance at the Future

“…Cole,Whitney, and Holland (1971) found support for this structure. Jackson, Holden, Locklin, and Marks (1984) suggested that a 5-factor model be used, based on cluster analyses of over 10,000 students who completed the Jackson Vocational Interest Survey. The US.…”

An Integrative Model of Factors Related to Computing Course Performance

“…Additionally, males and females are often considered as falling towards the higher and lower ends of this continuum respectively [23,24]. So, granted that programming is clustered with sciences in terms of its perceived attributes [25], that computer science students' vocational interests are similar to those of science and engineering students [26], and that science attitudes are positively related to computer attitudes [18,27], the present theoretical model included an indirect path linking gender and computer engagement by way of differences in tough-poise. The existence or otherwise of a bivariate gender -tough poise relationship bore upon the issue of whether females opting to specialize in programming-oriented computing exhibit more masculine personality characteristics than females generally.…”