The Rey AVLT Serial Position Effect: A Useful Indicator of Symptom Exaggeration?

Abstract

This investigation explored the usefulness of serial position patterns during word recall on the Rey Auditory Verbal Learning Test (RAVLT; Rey, 1964) as an indicator of poor effort. Significantly better recall for early (primacy) and recent (recency) material defines the serial position effect (SPE; Rundus, 1971). The SPE on the RAVLT was examined in four groups: normal controls (NC), symptom-coached simulators (SC), test-coached simulators (TC), and a group of moderate to severe subacute traumatic brain injury (TBI) patients. Normal control participants and TBI patients demonstrated the expected SPE. Only the SC simulators clearly suppressed the primacy effect. The SPE appears neither sensitive nor specific enough to be used independently of more sensitive symptom validity tests in the detection of suboptimal effort. It may be especially problematic when used with clients presenting with sophisticated styles of exaggeration and in settings with lower base rates of compromised effort.

Notes

*Portions of this study were presented at the 23rd Annual Conference of the National Academy of Neuropsychology, Dallas, Texas, October, 2003. The authors gratefully acknowledge Angie Capps and Nicole Salazar for their assistance with data collection, and Donna LaVoie for her helpful comments throughout the design and implementation of this work.

Note. Dissimilar subscripts within rows indicate significant group differences (p < .001).

Note. Dissimilar subscripts represent significant mean differences (p ≤ .01).

Note. Sensitivity = 61.5%; specificity = 66.7%; positive predictive value = 76.2%; negative predictive value = 50%.

Note. Sensitivity = 7.7%; specificity = 96.6%; positive predictive value = 80%; negative predictive value = 37.7%.

Note. Sensitivity = 59.6%; specificity = 73.3%; positive predictive value = 79.5%; negative predictive value = 48.8%.

Note. Sensitivity = 7.7%; specificity = 96.6%; positive predictive value = 80%; negative predictive value = 37.7%.

Note. PS cutoff = Primacy Sum Index cut score; P/L cutoff = Primacy/Learning Index cut score; SEN = sensitivity; SPEC = specificity; PPV = positive predictive value; NPV = negative predictive value; OCA = overall classification accuracy; C.I. = 95% confidence interval.

^aBase rate of symptom exaggeration supported by Mittenberg et al. (2002) and Larrabee (2003) for mild head injury compensation seekers.

^bBase rate of symptom exaggeration supported by Langeluddecke and Lucas (2003) for mild head injury compensation seekers.