C-66Recognition Discriminability (‘d): An Efficient Replacment for Forced-Choice Recogntion on the California Verbal Learning Test II (CVLT-II)
Lisa M. Isaac, Kenneth Stoler, Teresa G. Wilcox, R Johnson, Thomas Swirsky-Sacchetti · Archives of Clinical Neuropsychology · 2017
Objective: Performance validity tests (PVTs) are commonly used in neuropsychological evaluations (Meyers & Volbrecht, 2002). However, administration is subject to several constraints, including time and interference. The California Verbal Learning Test, Second Edition (CVLT-II), features a delayed Forced Choice Recognition (FC) as an embedded PVT. The FC is preceded by a Cued Recognition trial, which includes a Total Recognition Discriminability (d’) statistic. It is hypothesized that d’ would have acceptable classification accuracy and serve as an efficient PVT independent of the FC. Method: The sample consisted of 895 individuals who completed neuropsychological evaluation in a private practice setting. The sample was equally distributed among sex (Male = 50.17%) and was reasonably ethnically diverse (Caucasian = 62%). Each participant was administered the CVLT-II as a part of a larger neuropsychological battery. Results: Total Recognition Discriminability (d’) had excellent classification accuracy, [AUC] = .84, SE = .02, p < .001,. The suggested/cut score threshold of d’ = 2.6 resulted in strong specificity (.91) and moderate sensitivity (.61). Conclusion: These findings indicate preliminary evidence that Total Recognition Discriminability can serve as an effective replacement for Forced Choice Recognition. Further research on the correspondence of d’ to other performance validity tests is warranted.