Journal of Counseling Psychology 1991, Vol. 38, No. 3,350-366 Copyright 1991 by the American Psychological Association. Inc. 0022-0167/91/$3.00 Computer Versus Counselor Interpretation of Interest Inventories: The Case of the Self-Directed Search Itamar Gati and Dani Blumberg Hebrew University of Jerusalem, Jerusalem, Israel Interpretations of 100 career counselees' responses to the Self-Directed Search (SDS) were examined. The agreement between the scales (types) that were identified as relevant by the algorithm developed for this study and by 5 career counselors was as high as the agreement among the counselors (85%). The agreement between the counselors and Holland's 3-letter codes was lower (76%). In the 2nd study, insignificant correlations (r = .06 and -.03) were found between the counselors' judgments of the counselee's degree of interest crystallization and Holland's (1985a) measure of consistency, but their judgments were correlated (r = .41 and .45) with similarity within profile (an index of crystallization, Gati, 1985). In the 3rd study, various measures of the compatibility among the SDS's three major sections were found to be correlated with the counselor's judgments (.39 < r < .71). Implications for the interpretation of interest inventories are discussed. Interests constitute one of a family of constructs elaborated by psychologists to account for that part of the variance in occupational performance and satisfaction that cannot be explained by ability variables (Dawis, 1980). Hence, voca- tional interests are among the more important factors to be taken into account in career decisions (Holland, 1985a). The Self-Directed Search (SDS; Holland, 1985b) is one of the most widely used interest inventories (Krieshok, 1987). It is based on Holland's categorization of personality types and of environments into the following six categories: Realistic (R), Investigative (I), Artistic (A), Social (S), Enterprising (E), and Conventional (C). In the interpretation of the SDS, Holland refers Jo three aspects: (a) the counselee's main types or fields, (b) the consistency of the two types or fields with the highest ranking, and (c) the degree of differentiation between the types or fields. An additional aspect that may be added to these three is the compatibility or the coherence of the response patterns among the three major sections of the inventory: Activities, Competencies, and Occupations. The way in which career counselors interpret a counselee's pattern of responses in the SDS has significant implications for the counseling process and hence for the counselee. Counselors use various methods to interpret a counselee's interests as reflected in the SDS. Because the interpretations often do not coincide, the reliability of the interpretation and the degree of agreement between career counselors deserve attention. Some concepts and measures proposed by Holland (1985a) to enhance the interpretation of counselees' interests have theoretical and empirical limitations (Gati, 1985), and attempts have been made to improve or replace them (e.g., Gati, 1985, 1987). The widespread use of interest inventories that are based on Holland's (1973) typology, and in particular the recent computerized version of the SDS (Reardon, 1987), justifies not only efforts to examine the quality of interpreta- tions with Holland's concepts and the respective measures but also attempts to improve them. In the present research, we tested procedures that attempt to examine, formalize, and improve the interpretation of an individual's interest profile in the Self-Directed Search. We report three studies. In the first study, we reviewed problems associated with the three-letter interest codes and tested the possibility of identifying the relevant or prominent interest fields in the SDS by an algorithm that identifies a variable number of relevant fields. Then we compared career coun- selors' judgments to the outcomes of the algorithm and to Holland's proposal of using the three highest scales. In the second study, we compared Holland's consistency measure with an alternative definition and a respective measure for interest crystallization (Gati, 1985). We compared career counselors' judgments with the proposed measure and with Holland's consistency. In the third study, we examined var- ious measures for assessing the compatibility or coherence of the three major sections in the SDS and compared the agree- ment of these measures with career counselors'judgments. In the last section of the article, we discuss the implications of the results for the interpretation of interest inventories by career counselors and by computers. This research was supported by a grant from the Israel Foundations Trustees. We thank the five career counselors who served as expert judges and Tali Avishay, Sorel Cahan, Michal Givon, Elchanan I. Meir, Micha Wartski, and Dan Zakay for their helpful comments on an earlier version of this article. Correspondence concerning this article should be addressed to Itamar Gati, Department of Psychology, Hebrew University of Jeru- salem, Jerusalem 91905, Israel. General Information About the Method The three studies in this article have a number of common characteristics, which are described in this section to provide information about the SDS, the counselees whose SDS scores were analyzed, the career counselors who served as expert judges, and the procedure. 350 This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.