Analog validation of German-language symptom validity tests and the influence of coaching

Archives of Clinical Neuropsychology 20 (2005) 719–726 Analog validation of German-language symptom validity tests and the influence of coaching夽 Thomas Merten a,∗ , Paul Green b , Matthias Henry c , Nina Blaskewitz c , Robbi Brockhaus d a Vivantes Netzwerk für Gesundheit, Klinikum im Friedrichshain, Klinik für Neurologie, Landsberger Allee 49, D-10249 Berlin, Germany b Neurobehavioral Associates, Edmonton, Canada c Humboldt University, Berlin, Germany d Alexianer Krankenhaus, Krefeld, Germany Accepted 9 April 2005 Abstract Although symptom validity testing is an integral part of the repertory of neuropsychologists in a number of countries, this is not yet true for Germany. The German adaptations of two effort tests, the Medical Symptom Validity Test (MSVT) by Green and the Amsterdam Short-Term Memory Test (ASTM) by Schmand et al., were investigated with a German-language sample. An analog study was performed with 18 healthy experimental malingerers and 18 controls with a mean age of 25.4 years. The scenario contained detailed information about mild post-traumatic cognitive impairment, as well as an explicit warning against symptom exaggeration. In addition to MSVT and ASTM, the Trail Making Test (TMT), the Complex Figure Test (CFT), and Digit Span were performed. Half of the sample were also given Rey’s 15-Item-Test (FIT). Both groups were significantly different in all effort and performance measures, with the exception of the ratio TMT-B:TMT-A. With MSVT and ASTM, correct classification of group membership was between 97 and 100%. For the ratio TMT-B:TMT-A, there was a considerable overlap in the test scores for the two groups and the sensitivity of the FIT was too low. Although the ASTM and the MSVT were identified by a number 夽 The results of this study were presented as a poster at the 19th Annual Meeting of the German Neuropsychological Society (GNP) in Munich, September 2004. A more detailed report on the study as well as a translation of the scenario and the questionnaires used can be obtained from the authors. ∗ Corresponding author. Tel.: +49 3042211663. E-mail address: (T. Merten). 0887-6177/$ – see front matter © 2005 National Academy of Neuropsychology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.acn.2005.04.004 720 T. Merten et al. / Archives of Clinical Neuropsychology 20 (2005) 719–726 of subjects as possible effort measures, both tests obtained very good results within this analog design. © 2005 National Academy of Neuropsychology. Published by Elsevier Ltd. All rights reserved. Keywords: Malingering; Symptom validity testing; Neuropsychological assessment; Negative response bias; Memory 1. Introduction Psychologists have always made comments in reports about the effort that patients apply to testing, acknowledging that it is a significant variable, but objective testing of effort is a recent development. It is now standard practice in the USA and Canada for neuropsychologists to use effort testing routinely, especially if financial compensation is an issue but also more generally because poor effort will usually go undetected if it is not measured. In contrast to the remarkable interest that malingering research has received in North America, effort testing has had little impact so far on European countries. A major exception is the Amsterdam Short-Term Memory Test (Schagen, Schmand, de Sterke, & Lindeboom, 1997; Schmand et al., 1998), which has been adapted into English and German from the original Dutch. Also, the oral and computerized versions of Green’s Word Memory Test (WMT; Green, 2003) are available in German, Spanish, Dutch, French, English, Danish, and Turkish. In German-speaking countries, until now there have been a very small number of relevant publications on the subject of suboptimal performance. Also, symptom validity tests (SVTs) have not yet been employed by clinicians on a widespread basis. There have been two Germanlanguage studies (Brockhaus & Merten, 2004) using the WMT, one with patients with mental retardation, and the other with experimental malingerers and normal controls. The first study showed that the WMT was able to identify high performance motivation in almost all the patients with mental retardation. Only one participant was incorrectly classified. The second study compared performance of a healthy group of experimental malingerers (n = 100) with a healthy control group after standard test instructions (n = 27). The WMT was able to classify cases of good effort versus cognitive exaggeration with 100% accuracy in the study, replicating the finding of 100% correct classification by the WMT in an independent simulator study in English (Tan, Slick, Strauss, & Hultsch, 2002). In a more recent study, which is one of the few empirical studies on symptom validity testing ever published in the German language, Merten, Henry, and Hilsabeck (2004) used an analog design with a small sample size to investigate the applicability of German adaptations of foreign-language tests. In that study, three SVTs were used: the Word Memory Test, the Amsterdam Short-Term Memory Test, and the Word Completion Memory Test. For the SVTs, considerable effect sizes and a correct group classification of 95–100% was obtained. Although ratings of test difficulty given by all participants revealed that the SVT tasks were judged to be easier than the performance tests, analogue malingerers were unable to simulate the response patterns of truly brain-injured patients on the SVT measures. The present study was conceived to follow that line of research with a different set of tests and a different scenario, which included an elaborate warning that effort measurement was T. Merten et al. / Archives of Clinical Neuropsychology 20 (2005) 719–726 721 to be part of the assessment procedure and that overplaying the role would result in a loss of credibility in the overall pattern of performance. Moreover, a new test, the MSVT (Medical Symptom Validity Test; Green, 2004) was to be tested in the context of an analog design. The main question of the study was how well German test adaptations would work. One additional question to be addressed by the study was how well the participants would be able to identify the real symptom validity tests among a set of instruments, after receiving explicit warnings about effort assessment. To this effect, another test, Rey’s 15-Item-Test (FIT; Rey, 1958), was given to half of the experimental group and to half of the controls to see whether this test would be recognized as an effort measure very easily and, thereby, deflect attention from the other SVTs. After the experiment, the subjects were asked to identify possible effort measures. The hypothesis was that those participants who were given the FIT would easily recognize it as an SVT and, as a consequence, identify the other effort tests with less frequency. 2. Method 2.1. Research participants Thirty-six healthy, well-educated fluent sp (...truncated)