Comparison of various RAVLT scores in the detection of noncredible memory performance

• Published in: Archives of clinical neuropsychology Vol. 20; no. 3; pp. 301 - 319
• Main Authors: Boone, Kyle Brauer, Lu, Po, Wen, Johnny
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Ltd 01.05.2005; Elsevier
• Subjects: Adolescent; Adult; Aged; Biological and medical sciences; Cognition Disorders - diagnosis; Cognition Disorders - psychology; Disability Evaluation; Educational Status; Female; Humans; Male; Malingering; Malingering - diagnosis; Medical sciences; Memory - physiology; Middle Aged; Motivation; Neuropsychological Tests; Psychology. Psychoanalysis. Psychiatry; Psychometrics; Psychometrics. Diagnostic aid systems; Psychopathology. Psychiatry; RAVLT; Rey Auditory Verbal Learning Test; Sensitivity and Specificity; Suspect effort; Techniques and methods; Suspect effort; Rey Auditory Verbal Learning Test; RAVLT; Malingering; Human; Memory disorder; Verbal learning; Cognitive disorder; Effort; Neuropsychological test; Psychometrics; Verbal memory; Auditory Verbal Learning Test Rey; Response falsification; Simulation
• ISSN: 0887-6177; 1873-5843
• DOI: 10.1016/j.acn.2004.08.001

Sixty-one noncredible patients (as documented by psychometric and behavioral criteria) scored significantly below 25 controls and 88 credible clinic patients with no motive to feign on most RAVLT scores, including added indices of implicit and “automatic” memory. A combination of true recognition (i.e., recognition minus false positives) + implicit memory score (i.e., the number of word stems completed with RAVLT items) + “automatic” memory score (i.e., the number of correct temporal order judgments) ≤22 was associated with 75.7% sensitivity with specificity at 91.5%. However, sensitivity was nearly as high when scores available from the standard RAVLT administration alone (i.e., no word stem or temporal order trials) were combined. Specifically, a cut-off of ≤12 for true recognition (recognition minus false positives) + primacy recognition (i.e., number of words recognized from the first third of the test) was associated with 73.8% sensitivity at 90% specificity. These results indicate that combined indices of recognition memory from the RAVLT are effective in identifying noncredible memory performance in “real world” samples and are modestly superior to the 67.2% sensitivity obtained with the standard recognition score.