An investigation into the comparative validity of several diagnostic tests and test batteries

Abstract

existence of a relatively small number of modal types (as defined by the major CPI factors) suggests that code type approaches may be practicable with the CPI. Nontest characteristics of such groups could be noted and interpretative aids for the CPI prepared. Of course, the results of this study should be evaluated within the context of the sample involved. It remains to be seen whether other samples (psychiatric patients, groups that have professed adjustment problems, etc.) exhibit the same pattern frequencies as normals. In any case, the feasibility of a type approach with the CPI seems promising enough to encourage future inquiries. SUMMARYCPI scale scores were correlated with specially-scored CPI factor scales, and the results indicated that such factor scores could be estimated reasonably by the standard scales. By the use of CPI scales as measures of the two major factors of that instrument, the incidence of profile types in a normal sample was tabulated. For both males and females, six types accounted for SOY0 of the protocols: the first three types accounted for over 50% of the individuals. It was suggested that this type of analysis might contribute to interpretation of CPI protocols. MARKS, P. A, and SEEMAN, W. The Actuarial Description ofA b n m l Personality. Balitrnore: 5. MEOARGEE, E. I. The California Psychological Invent0 y Handbook. San Francisco: Jossey-Bass, 6. MITCHELL, J. V., JR. and PIERCE-JONES, J. A factor analysis of Gough's California Psychologi-7. NICHOLS, R. C. and SCHNELL, R. R. Factor scales for the California Psychological Inventory.

The Scientific Status of Projective Techniques

“…Finally, there are few data bearing on the incremental validity of human figure drawings above and beyond either psychometric or demographic information. Wildman and Wildman (1975) found that adding the H-T-P to the Bender-Gestalt figure drawing test decreased the accuracy of clinicians' classifications (of individuals as either psychiatric patients or nurses) from 62% to 53%. We are unaware of any studies demonstrating that human figure drawings offer psychologically useful information over and above the MMPI-2, psychiatric interviews, demographic data, or other information that is often readily available in clinical settings.…”

“…In contrast, adding the TAT to both the Rorschach and MMPI led to essentially no increases in validity. Wildman and Wildman (1975) asked a group of clinical psychologists to determine, on the basis of various combinations of test results, whether respondents were psychiatric patients or nonpatients (nurses) who had been matched to the patients for age and education. Although adding the MMPI to the TAT resulted in an increase in accuracy from 57% to 80%, adding the TAT to the MMPI resulted in a decrease in accuracy from 88% to 80%.…”

Human figure drawing in adults: An update of the empirical evidence, 1967–1982.

“…For example, Strieker (1967) asked judges to decide whether protocols from adolescents were from normals or psychiatric patients, and obtained 66 to 73% correct judgements, compared to a base rate of 57% if all drawings had been classified pathological. Most classification attempts have been largely unsuccessful (Rubin, Ragins, Schachter, & Wimberly, 1979;Wanderer, 1969;Wildman & Wildman, 1975), or have provided a degree of improvement above chance that, though statistically significant, is so slight as to be meaningless (Cressen, 1975;Strieker, 1967;Watson, 1967a). Experienced judges have proved no more accurate than inexperienced ones; in fact, Strieker (1967) found that experienced clinicians were less accurate than student judges, possibly because they were reluctant to abandon their own individualistic diagnostic criteria in favour of the more valid written guidelines that were given to all judges.…”

Practical Considerations In The Assessment Of Potential Abilities In Applicants For Benefit Programs

“…Briefly, a ratio is calculated by dividing the number of the 20 empirically-determined easiest items on the instrument (1,2,3,12,15,20,21,23,24,28,31,32,33,34,35,36,37,38,39,40) by the total number of matching items with incorrect answers. Predicting that patients scoring .125 and higher on this index are trying to hide their true level of intelligence, while lower-scorers are making a good-faith effort produces a "hit-rate" (Wildman & Wildman, 1975) of 77%. Specifically, this index correctly identifies 87% of those downplaying their abilities and 67% of those responding honestly, employing here a broad range of data sources outside the NBCAI itself, suggesting more false positives than false negatives with respect to having the condition of malingering.…”