Detection of Malingering and Invalid Test Scores

Archives of Clinical Neuropsychology, Vol. 14, No. 6, pp. 563–564, 1999 Copyright © 1999 National Academy of Neuropsychology Printed in the USA. All rights reserved 0887-6177/99 $–see front matter PII S0887-6177(98)00155-3 Book and Test Review Detection of Malingering and Invalid Test Scores, by Ralph M. Reitan and Deborah Wolfson. Tucson, AZ: Neuropsychology Press, 1997, 164 pp. With considerable interest, I read Reitan and Wolfson’s new book on detection of malingering and invalid test scores. As with all of their books, this one is well-conceptualized, well-organized, clearly written, well-referenced, and illustrated by a wealth of case study materials to illustrate the principles and/or procedures being discussed. The book begins with a presentation of the issue of malingering and the difficulty of determining that malingering has really occurred. The authors consider the difficulties of trying to study this phenomenon by using college students or other “normal” subjects who are simply told to feign a neurobehavioral syndrome, such as traumatic brain injury. As they indicate, it is not possible to assume that normal subjects told to “fake bad,” after being told about symptoms of brain injury, will actually be able to simulate the neuropsychological impairments caused by brain damage. As an alternative approach, these investigators suggest that a different model of study is one in which a person is tested on two or more occasions, using the same tests, and then the consistency of responding over the repeated testing sessions could be used as the dependent variable. Since, barring late complications, individuals with nonprogressive neurobehavioral syndromes should perform similarly or better on repeated test administrations, an index of invalid test results or malingering can be obtained based on poorer performance on the subsequent testing. The major drawback to this approach is that it requires two sets of the critical data, in this case including scores from core tests of the Halstead Reitan Neuropsychological Test Battery and the Wechsler Adult Intelligence Scale (WAIS) or WAIS-R. Since there is great variability among the tests that neuropsychologists employ in making their conclusions, obtaining two successive sets of the required test scores can obviously be very problematic. There is relevant case law bearing on the issue of test selection. In Chapple v. Ganger (1994), a federal district court ruled that the fixed Halstead Reitan Neuropsychological Test Battery met the Daubert standard for the admissibility of scientific evidence. Data obtained using a flexible neuropsychological assessment approach did not. For a review of this case, see Reed (1996). Three indexes to evaluate validity of responding are derived by Reitan and Wolfson using the test-retest procedure. These include the following: The Retest Consistency Index is calculated from comparative scores on the comprehension, picture arrangement, and digit symbol subtests from the Wechsler scales, as well as the Category Test, Trails B, and the Tactual Performance Test Localization scores from the Halstead-Reitan battery. The Response Consistency Index is derived from comparative scores on the information, comprehension, arithmetic, similarities, and vocabulary subtests from the Wechsler scales. Finally, the Dissimulation Index is based on the Retest Consistency Index 563 564 Book and Test Review and the Response Consistency Index, and it is obtained by adding the scores for the other two Indexes. Reitan and Wolfson take the reader through the simple steps of calculating the various indexes, and then they provide cutoff scores that appear to reliably separate valid from invalid test results. The Indexes are based on a comparative study between two groups of potentially braininjured individuals, one group was involved in litigation to recover damages allegedly resulting from a brain injury, while the other group was not (nonlitigation group). Each group had 20 subjects, presumably selected from the authors’ extensive archives. The authors indicate that the nonlitigants, as a group, were more severely injured than the litigants. Frequency and duration of loss of consciousness was greater in the nonlitigants. The nonlitigants were younger by an average of approximately 12 years. Education level was comparable in the two groups (12.30 years for the nonlitigants and 13.95 years for the litigants). The average time between the two assessments was comparable (12.00 months for the nonlitigants and 14.45 months for the litigants). Whether the differences between the two groups influenced the data needs to be evaluated by additional research. Neither group had participated in cognitive rehabilitation in a systematic way. Reitan and Wolfson stress the need for additional research to gain an understanding of the characteristics of individuals who produce invalid test results, as compared to those who produce valid scores. They also point out a need for investigators to study the influence of examiners in producing valid versus invalid test results. I am sure that there will be a great deal of research on these indexes, now that they are available. It will be helpful to compare findings from these indexes with more widely used forced-choice testing procedures. The Retest Consistency Index, the Response Consistency Index, and the Dissimulation Index provide us with another way to assess the possibility of malingering. These procedures are interesting because they provide objective measures of validity utilizing scores from neuropsychological tests. I find them appealing because they eliminate the boredom for the examiner and the examinee of taking a forced-choice test, a procedure that often makes the results of forced-choice testing inconclusive. Neuropsychologists will be well-advised to use this new methodology, supplementing it with other measures as needed. Detection of malingering and invalid responding is difficult, and the basis for what seems to be a malingered test profile may not always be apparent. Cultural factors that often affect test performance can result in what appear to be invalid profiles, but the basis for the findings may be the approach and world view that one brings to the testing environment rather than a deliberate attempt to magnify symptoms for secondary gain. REFERENCES Chapple v. Ganger. (1994). 851 F. Supp. 1481 (E. D. Wash, 1994). Reed, J. E. (1996). Fixed vs. flexible neuropsychological test batteries under the Daubert standard for the admissibility of scientific evidence. Behavioural Sciences and the Law, 14, 315–322. Thomas L. Bennett Brain Injury Recovery Program Fort Collins, CO 80524  (...truncated)