Journal of Counseling Psychology 1991, Vol. 38, No. 3, 331-341 Copyright 1991 by the American Psychological Association, Inc. 0022-0167/91/S3.00 Development and Validation of the Scientist-Practitioner Inventory for Psychology Frederick T. L. Leong and Peter Zachar Southern Illinois University at Carbondale This article reports 3 studies constituting the development and validation of the Scientist- Practitioner Inventory (SPI), an instrument designed to measure career specialty interests of psychology students. In the 1st study, evidence for the factorial validity of the scales, test-retest reliability, freedom from response-set biases, and construct validity are presented. The 2nd study presents cross-validation evidence of a 2nd-order factor structure, internal consistency estimates, and construct validity that is similar to that found in the first study. Also presented is evidence for the superior criterion-related concurrent validity of the SPI in comparison with Holland's (1985) Vocational Preference Inventory in predicting a selected criterion. The 3rd study presents further cross-validating evidence and indications of the SPI's ability to reflect different patterns of interests depending on the nature of the sample. Implications for use of the instruments in career counseling, psychology of science, and organizational research are discussed. Vocational psychologists have long attempted to show that different careers attract different kinds of people. Roe (1953) demonstrated that physical scientists have characteristics that are different from those of both social scientists and artists. On the basis of this formulation, career counselors have made an extended effort to help clients choose occupations that match their personalities. Besides choosing global occupational fields (labeled inter- occupational choice by Winer, 1981), many people are re- quired to choose particular specializations (intraoccupational choice; Winer, 1981). Medical students can become neurolo- gists, pediatricians, cardiologists, general practitioners, or other specialists. Law students may choose to specialize in criminal law, corporate law, environmental law, and so forth. In many occupations specialization choices cannot be avoided. Because different subspecialties supposedly attract different kinds of people, most career counselors work under a tacit assumption that specialty choice follows naturally once a basic occupational field has been decided upon. However, for many occupations there is no reason to assume that choice of specialty is an easier choice, or a less important choice, than choice of career field. Despite Ginzberg, Ginsburg, Ax- elrad, and Henna's (1951) formulation of specification as an important component of career development, vocational psy- chology has focused primarily on factors associated with the initial career or occupational choice, whereas career specialty choice has been a relatively neglected area of research. The empirical literature on differences among psychologists (Coan, 1979; Garfield & Kurtz, 1976; Gottfredson, 1987; We thank Howard E. A. Tinsley and Sandra Leong for their assistance with this project. Peter Zachar is currently a psychology intern at the Counseling Center, University of Iowa, Iowa City, Iowa 52242. Correspondence concerning this article should be addressed to Frederick T. L. Leong, who is now at the Department of Psychology, Ohio State University, 142 Townshend Hall, Columbus, Ohio 43210. Kimble, 1984;Lockman, 1964; Richards, 1974; Shaffer, 1953; Youniss, Loir, & Stefic, 1985) has consistently found dimen- sions that have variously been referred to as an objective- versus-intuitive orientation, an objective-versus-subjective orientation, or a scientist-versus-practitioner orientation, in- dicating that the scientist-practitioner split is one of the most important constructs for explaining differences among psychologists. Counseling psychologists have put extended effort into formulating a viable approach to training psychologists within a scientist-practitioner framework, which is most often re- ferred to as the Boulder model. For example, Gelso (1979) proposed nine criteria for graduate programs geared toward training therapists to be scientists. The relative merits of the implementation of these criteria were empirically investigated by Royalty, Gelso, Mallinckrodt, & Garret (1986). Magoon and Holland (1984) and Holland (1986) argued that scientists and practitioners have different personality styles, which are best suited to different work environments. Wampold (1986) argued that the assumption that counseling students are purely social creatures with natural antipathy toward doing statistics is false. Howard (1986) offered a compromise solution by invoking a Thomistic-Aristotelian paradigm of causality for scientists that is potentially more congenial to practitioner world views than is the reductionistic-positivistic causal par- adigm favored by 20th-century American psychology. The literature on training psychologists within a scientist- practitioner framework indicates that the shared view oper- ating in the field is the following: (a) There is something very different about scientific training in comparison with practi- tioner training; (b) some people are more inclined toward science, whereas other people are more inclined toward prac- tice; (c) those inclined toward practice generally do not like science, and presumably those inclined toward science would not like practice. In spite of this consensus, there has been relatively little direct empirical research on scientist and prac- titioner orientations. 331