Search

The purpose of this study was to determine how well scores from the Rey Complex Figure Test (RCFT) could serve as embedded measures of performance validity in a large, heterogeneous clinical sample at an urban-based Veterans' Affairs hospital. Participants were divided into credible performance (n = 244) and noncredible performance (n = 87) groups based on common performance validity tests during their respective clinical evaluations. We evaluated how well preselected RCFT scores could discriminate between the 2 groups using cut scores from single indexes as well as multivariate logistic regression prediction models. Additionally, we evaluated how well memory error patterns (MEPs) could discriminate between the 2 groups. Optimal discrimination occurred when indexes from the Copy and Recognition trials were simultaneous predictors in logistic regression models, with 91% specificity and at least 53% sensitivity. Logistic regression yielded superior discrimination compared with individual indexes and compared with the use of MEPs. Specific scores on the RCFT, including the Copy and Recognition trials, can serve as adequate indexes of performance validity, when using both cut scores and logistic regression prediction models. We provide logistic regression equations that can be applied in similar clinical settings to assist in determining performance validity.

Publication of the third editions of the Wechsler intelligence and memory batteries in 1997 created a need for research identifying Wechsler Adult Intelligence Scale-Third Edition/Wechsler Memory Scale-Third Edition (WAIS-III/WMS-III) profile patterns associated with neuropathology. The WAIS-III/WMS-III Technical Manual offers data on various diagnostic groups, including traumatic brain injury (TBI). Hawkins (1998) employed Technical Manual data to propose certain diagnostic guidelines. In order to validate the conclusions put forth by Hawkins as they apply to brain injury, we examined WAIS-III and WMS-III profiles in an independent sample of 46 TBI cases. As expected, the WAIS-III Processing Speed Index (PSI) was more sensitive to brain injury than other WAIS-III composites; and specific WAIS-III scores were stronger than certain WMS-III scores. On the other hand, the predicted relationship for WMS-III auditory and visual indexes was not found. The lack of specificity for TBI of the proposed index comparisons confirms the need to validate such hypotheses in independent samples.

WAIS-III profile interpretation typically involves the analysis of disparities between summary and subtest scores. Determination of clinically meaningful differences has been guided by published rates of discrepancies based upon the normative sample. However, noting that a particular profile discrepancy is uncommon among normals only represents one step in the interpretive process. Rates of such pairwise differences have yet to be published for clinical samples. In order to begin filling this void in the literature, we present frequencies of composite and subtest differences from a postacute traumatic brain injury sample. The practical utility of these tables, in conjunction with the normative tables, is demonstrated.