The current study attempted to improve upon the efficiency and accuracy of one of the most frequently administered measures of test validity, the Test of Memory Malingering (TOMM) by utilizing two short forms (TOMM trial 1 or TOMM1; and errors on the first 10 items of TOMM1 or TOMMe10). In addition, we cross-validated the accuracy of five embedded measures frequently used in malingering research. TOMM1 and TOMMe10 were highly accurate in predicting test validity (area under the curve [AUC]=92% and 87%, respectively; TOMM1≤40 and TOMMe10≥1; sensitivities>70% and specificities>90%). A logistic regression of five embedded measures showed better accuracy compared with any individual embedded measure alone or in combination (AUC=87%). TOMM1 and TOMMe10 provide evidence of greater sensitivity to invalid test performance compared with the standard TOMM administration and the use of regression improved the accuracy of the five embedded cognitive measures.
Embedded validity measures support comprehensive assessment of performance validity. The purpose of this study was to evaluate the accuracy of individual embedded measures and to reduce them to the most efficient combination. The sample included 212 postdeployment veterans (average age = 35 years, average education = 14 years). Thirty embedded measures were initially identified as predictors of Green's Word Memory Test (WMT) and were derived from the California Verbal Learning Test-Second Edition (CVLT-II), Conners' Continuous Performance Test-Second Edition (CPT-II), Trail Making Test, Stroop, Wisconsin Card Sorting Test-64, the Wechsler Adult Intelligence Scale-Third Edition Letter-Number Sequencing, Rey Complex Figure Test (RCFT), Brief Visuospatial Memory Test-Revised, and the Finger Tapping Test. Eight nonoverlapping measures with the highest area-under-the-curve (AUC) values were retained for entry into a logistic regression analysis. Embedded measure accuracy was also compared to cutoffs found in the existing literature. Twenty-one percent of the sample failed the WMT. Previously developed cutoffs for individual measures showed poor sensitivity (SN) in the current sample except for the CPT-II (Total Errors, SN = .41). The CVLT-II (Trials 1-5 Total) showed the best overall accuracy (AUC = .80). After redundant measures were statistically eliminated, the model included the RCFT (Recognition True Positives), CPT-II (Total Errors), and CVLT-II (Trials 1-5 Total) and increased overall accuracy compared with the CVLT-II alone (AUC = .87). The combination of just 3 measures from the CPT-II, CVLT-II, and RCFT was the most accurate/efficient in predicting WMT performance.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.