Loading…

A Factor-Analytic Study of the Construct Validity of Holland's Self-Directed Search Test

A random sample of 2626 Canadian professional accountants was mailed a questionnaire containing the Self-Directed Search instrument (SDS) developed by Holland. The 1206 completed questionnaires were used to obtain more evidence about the construct validity of the SDS. A method of common factor analy...

Full description

Saved in:
Bibliographic Details
Published in:Educational and psychological measurement 1981-07, Vol.41 (2), p.425-437
Main Authors: Rachman, D., Amernic, J., Aranya, N.
Format: Article
Language:English
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A random sample of 2626 Canadian professional accountants was mailed a questionnaire containing the Self-Directed Search instrument (SDS) developed by Holland. The 1206 completed questionnaires were used to obtain more evidence about the construct validity of the SDS. A method of common factor analysis was used. In general, it was found that the structure of the SDS instrument as a whole is quite clear and that most of the items which are supposed to measure one personality type form a unidimensional set. The Activity and Ability subtests cannot discriminate between the Enterprising and Social types. Only the Interest subtest can differentiate among all the personality types prescribed by Holland. A confirmatory factor analysis provided support for the result that the SDS measures six factors: Realistic, Investigative, Artistic, Social-Enterprising (combined), Conventional, and a sixth general interest factor. Generally, in each subtest and for the entire test, the psychological relationship among types confirms the hexagon model proposed by Holland and others. Correlations between the scales of personality types based on the self-report and an index of social desirability were quite low. It is concluded that the theoretical structure of the hexagon model appears to be sound, but that the tools of measurement (i.e., the SDS subtests) may require improvement.
ISSN:0013-1644
1552-3888
DOI:10.1177/001316448104100221