Can face recognition be selectively preserved in some cases of amnesia? A cautionary tale

Evidence suggests that some patients with isolated hippocampal damage appear to present with selective preservation of unfamiliar face recognition relative to other kinds of visual test stimuli (e.g., words). Bird and Burgess (2008) formulated a review and secondary analysis of a group of 10 cases all tested on a clinical assessment of word and face recognition memory (RMT, Warrington, 1984), which confirmed the key memory dissociation at the group level. The current work provides an updated secondary analysis of such cases with a larger published sample (N = 52). In addition to group-level analyses, we also re-evaluate evidence using a single case statistical approach (Crawford & Garthwaite, 2005), enabling us to determine how many would make criteria for a 'classical dissociation' (Crawford, Garthwaite, & Gray, 2003). Overall, group-level analyses indicated the key pattern of significant differences confined to words was limited to small control sample comparisons. When using the large control sample provided by Bird and Burgess (2008), hippocampal cases as a group were significantly poorer for both classes of items. Furthermore, our single-case approach indicated few had a performance pattern of a relative difference across face > word categories that would meet statistical significance; namely within individual differences across categories that would warrant a significant 'classical dissociation'. Moreover, these analyses also found several cases with a 'classical dissociation' in the reverse direction: namely preserved recognition of words. Such analyses serve to demonstrate the need for a more conservative statistical approach to be undertaken when reporting selective 'preservation' of a category in recognition memory. Whilst material specificity has important implications for understanding the role of the hippocampus in memory, our results highlight the need for statistical methods to be unquestionably rigorous before any claims are made. Lastly, we highlight other methodological issues critical to group analyses and make suggestions for future work.

Towards an embedded symptom validity indicator for the rivermead postconcussion symptom questionnaire

This simulation study aimed to develop and test an embedded symptom validity indicator for use with the Rivermead Post Concussion Symptoms Questionnaire (RPQ). Seven mild traumatic brain injury (mTBI) experts assisted in generating a shortlist of 15 atypical postconcussion symptoms. A modified RPQ (mRPQ) was constructed by embedding these items with the 16 standard RPQ items. Eighty-four mTBI simulators completed the mRPQ after random allocation to a symptom-exaggeration condition (ES group; n = 46) or simulation-only condition (S group; n = 38). They also completed the Test of Memory Malingering (TOMM) and the Recognition Memory Test (RMT), and the groups were re-formed and compared using modified criteria for Malingered Neurocognitive Dysfunction (mMND). There was no significant group by item-type interaction for either grouping methods (i.e. as allocated or as reclassified). Exploratory analyses of the main effects showed that ES group had higher mRPQ scores than the S group, and standard symptoms were endorsed more than atypical symptoms. If further developed, the mRPQ could eventually aid the screening of PCS over-reporting.

Cross-validation of picture completion effort indices in personal injury litigants and disability claimants

Picture Completion (PC) indices from the Wechsler Adult Intelligence Scale, Third Edition, were investigated as performance validity indicators (PVIs) in a sample referred for independent neuropsychological examination. Participants from an archival database were included in the study if they were between the ages of 18 and 65 and were administered at least two PVIs. Effort measure performance yielded groups that passed all or failed one measure (Pass; n= 95) and failed two or more PVIs (Fail-2; n= 61). The Pass group performed better on PC than the Fail-2 group. PC cut scores were compared in differentiating Pass and Fail-2 groups. PC raw score of ≤12 showed the best classification accuracy in this sample correctly classifying 91% of Pass and 41% of Fail-2 cases. Overall, PC indices show good specificity and low sensitivity for exclusive use as PVIs, demonstrating promise for use as adjunctive embedded measures.

Detecting Response Bias with Performance Patterns on an Expanded Version of the Controlled Oral Word Association Test

The present study investigated whether speeded word generation performance patterns seen in healthy subjects are also produced in genuine and feigned traumatic brain injury (TBI). An expanded version of the Controlled Oral Word Association Test was administered to healthy controls, TBI patients, simulated malingerers, and probable clinical malingerers. Four performance patterns were operationalized. Three of these patterns were replicated in the healthy control sample and found to be unaltered by genuine TBI. They were then combined into a logistic regression model that discriminated well between examinees who put forth adequate effort and those who evidenced response bias.

The Performance of Individuals with Mental Retardation on Cognitive Tests Assessing Effort and Motivation

The purpose of this study was to determine which tests of effort and motivation would be appropriate for use with patients with mental retardation when feigning of cognitive deficits is suspected. The seven measures evaluated included the WMS-III Rarely Missed Index Test, forced-choice recognition portion of the California Verbal Learning Test-II, Reliable Digit Span test, Rey 15-Item Test, Rey Dot Counting Test, the Rey 15-Item Test with Recognition Trial, and the Vocabulary (V)-Digit Span (DS) difference score. Results indicated that the forced-choice portion of the CVLT-II, the V-DS difference score, and the Rarely Missed Index Test from the WMS-III might be appropriate for use with this population with passing rates of 89%, 98%, and 91% respectively.

Further Validation of the Expanded Auditory Verbal Learning Test for Detecting Poor Effort and Response Bias: Data from Temporal Lobectomy Candidates

The present study aimed to further examine the specificity of the Exaggeration Index of the expanded Auditory Verbal Learning Test procedure for detecting inadequate effort or response bias in a nonlitigating sample with unequivocal brain damage. Data from 56 patients with intractable epilepsy undergoing presurgical evaluation for temporal lobectomy were analyzed. In this sample, specificity was excellent at the recommended cut-off (94%), and comparable to that of another well-researched instrument for detecting inadequate effort, the Recognition Memory Test. Findings strongly support the generalizability of the specificity estimates obtained by Barrash et al. and, more broadly, the clinical utility of the Exaggeration Index. To facilitate diagnostic decision-making, the probability of invalid performances associated with each score on the Exaggeration Index, based on all clinical cross-validation samples reported to date, is also presented.

Detecting Feigned Impairment with the Digit Span and Vocabulary Subtests of the Wechsler Adult Intelligence Scale—Third Edition

The WAIS-III Digit Span and Vocabulary subtests were investigated as indicators of feigned cognitive impairment. Participants included 64 undergraduates randomly assigned to control, symptom-coached, or test-coached groups. Six previously researched validity indicators were examined. We hypothesized that symptom-coached participants would perform worse relative to test-coached simulators. Analyses determined both simulator groups performed lower than controls on all indicators except Vocabulary. Symptom-coached participants, however, did not differ from test-coached participants on any indicator. Classification accuracies for these six indicators ranged from 42 to 78%. While the WAIS-III validity indicators hold some promise, they should not be employed as independent measures.

Detecting Insufficient Effort Using the Seashore Rhythm and Speech-Sounds Perception Tests in Head Injury

This study examined the capacity of the Seashore Rhythm Test (SRT) and the Speech-Sounds Perception Test (SSPT) to detect insufficient effort in a clinical sample. Forty-six participants with financially compensable mild head injury who obtained scores indicative of insufficient effort on multiple measures were compared to 49 participants with brain injury who were not involved in litigation. Receiver operating characteristic (ROC) curve analysis indicated that both the SRT (AUC = .84) and SSPT (AUC = .80) were significant (p < .001) predictors of insufficient effort. Maximizing sensitivity and specificity, the optimal cutoff scores were 8 errors on the SRT and 10 errors on the SSPT. Combining both variables into a logistic regression function increased the diagnostic efficiency.