Quantified Bender-Gestalt protocols of 120 clients usin the Hutt, Pascal-Suttell, and Pauker scoring systems. Scores were correlate3 with two criteria of psychopathology, diagnosis of psychiatric disability and self-reported anxiety, and age and I&. The results suggest that: (a) Ratings of B-G protocols are highly reliable; (b) B-G scores are not predictive of global sychopathology; and (c) B-G scores are correlated moderately with age and fQ. This evidence does not support the validity of B-G scoring systems as indicators of psychopathology.
At least eight objective scoring systems for the Bender-Gestalt have been developed in the last 30 years, e.g., Billingslea, Hain, Hutt, Kitay, Koppitz, Pascal-Suttell, Pauker, and Peek-Quast. The indices that result from these scoring procedures purport to measure a variety of clinically relevant constructs, such as organicity, learning disabilities, ego-strength, and various types of psychopathology. Three of the scoring systems claim to assess psychopathology, as defined by psychiatric diagnosis (Hutt, 1977; Pascal & Suttell, 1951; Pauker, 1976), although the accumulated evidence that concerns their validity for this criterion is ambiguous (Tolor & Brannigan, 1980;Tolor & Schulberg, 1963). I n general, the empirical research literature with regard to the validity of B-G scoring systems is difficult to interpret because of the nonstandard administration of the test and the use of inappropriate or poorly defined criteria.
Furthermore, B-G scoring systems assume a unidimensional structure that underlies the diverse ratings of the protocols, i.e., the separate ratings of the characteristics and deviations of the designs are summed into a total index score. If this assumption is not warranted, then aggregating extraneous ratings together with the potentially useful diagnostic information may be another source of confusion in the validity literature. In fact, three extant studies of the dimensionality of B-G ratings did not produce consistent conclusions. The investigations by Haynes (1970) andHoffmann (1976) demonstrated that the Pascal-Suttell and Koppitz systems are multidimensional, while Stone (1966) concluded that the Pascal-Suttell is essentially unidimensional.
The purposes of this study were (a) to evaluate the validity of three B-G scoring systems (Hutt, Pascal-Suttell, and Pauker), each of which results in a total score that purports to differentiate between normals and various psychiatric diagnoses; and (b) to assess the dimensionality of B-G ratings across the three scoring systems and evaluate the validity of the resulting factors.
METHOD
Subjects
The research sample consisted of 120 persons referred to the Little Rock office of the Arkansas Rehabilitation Service. All candidates for rehabilitation services receive thorough medical and/or psychiatric examinations as a basis for determining eligibility for services. Each S was evaluated by a licensed psychological examiner, who administered the WAIS, Bender-Gestalt, Draw-A-Person, Sixteen Personality Factor Questionnaire (Form E) , and other standard tests and inventories.