Bridging the gap between neurocognitive processing theory and performance validity assessment among the cognitively impaired: a review and methodological approach

EM-COGLOAD: An investigation into age and cognitive load detection using eye tracking and deep learning

Alzheimer's Disease is the most prevalent neurodegenerative disease, and is a leading cause of disability among the elderly. Eye movement behaviour demonstrates potential as a non-invasive biomarker for Alzheimer's Disease, with changes detectable at an early stage after initial onset. This paper introduces a new publicly available dataset: EM-COGLOAD (available at https://osf.io/zjtdq/, DOI: 10.17605/OSF.IO/ZJTDQ). A dual-task paradigm was used to create effects of declined cognitive performance in 75 healthy adults as they carried out visual tracking tasks. Their eye movement was recorded, and time series classification of the extracted eye movement traces was explored using a range of deep learning techniques. The results of this showed that convolutional neural networks were able to achieve an accuracy of 87.5% when distinguishing between eye movement under low and high cognitive load, and 76% when distinguishing between the oldest and youngest age groups.

The Effects of Cognitive Load on Strategy Utilization in a Forced-Choice Recognition Memory Performance Validity Test

Abstract. Despite the importance of detecting feigned cognitive impairment, we have a limited understanding of the theoretical foundation of the phenomenon and the factors that affect it. Studies regarding the formation and implementation of feigning strategies during neuropsychological assessments are numbered, though there are indications that they tax cognitive resources. The current study assessed the effect of cognitive load manipulation on feigning strategies. To achieve this aim, we utilized a 2 × 2 experimental design; condition (simulators/honest responders) and cognitive load (load/no load) were manipulated while participants ( N = 154) performed a well-established performance validity test (PVT). The cognitive load manipulation reduced the quantity of feigning strategies, while also affecting their composition (i.e., strategies tended to be more intuitive). This suggests that reduced cognitive resources among those feigning cognitive impairment may impact the use of in-vivo feigning strategies. These findings, though preliminary, will hopefully encourage further research that will uncover the cognitive factors involved in the utilization of feigning strategies in neuropsychological assessments.

Flipping the Script: Measuring Both Performance Validity and Cognitive Ability with the Forced Choice Recognition Trial of the RCFT

In this study we attempted to replicate the classification accuracy of the newly introduced Forced Choice Recognition trial (FCR) of the Rey Complex Figure Test (RCFT) in a clinical sample. We administered the RCFT_FCR and the earlier Yes/No Recognition trial from the RCFT to 52 clinically referred patients as part of a comprehensive neuropsychological test battery and incentivized a separate control group of 83 university students to perform well on these measures. We then computed the classification accuracies of both measures against criterion performance validity tests (PVTs) and compared results between the two samples. At previously published validity cutoffs (≤16 & ≤17), the RCFT_FCR remained specific (.84-1.00) to psychometrically defined non-credible responding. Simultaneously, the RCFT_FCR was more sensitive to examinees' natural variability in visual-perceptual and verbal memory skills than the Yes/No Recognition trial. Even after being reduced to a seven-point scale (18-24) by the validity cutoffs, both RCFT recognition scores continued to provide clinically useful information on visual memory. This is the first study to validate the RCFT_FCR as a PVT in a clinical sample. Our data also support its use for measuring cognitive ability. Replication studies with more diverse samples and different criterion measures are still needed before large-scale clinical application of this scale.

Identifying Feigned Cognitive Impairment: Investigating the Utility of Diffusion Model Analyses

Forced-choice performance validity tests are routinely used for the detection of feigned cognitive impairment. The drift diffusion model deconstructs performance into distinct cognitive processes using accuracy and response time measures. It thereby offers a unique approach for gaining insight into examinees' speed-accuracy trade-offs and the cognitive processes that underlie their performance. The current study is the first to perform such analyses using a well-established forced-choice performance validity test. To achieve this aim, archival data of healthy participants, either simulating cognitive impairment in the Word Memory Test or performing it to the best of their ability, were analyzed using the EZ-diffusion model (N = 198). The groups differed in the three model parameters, with drift rate emerging as the best predictor of group membership. These findings provide initial evidence for the usefulness of the drift diffusion model in clarifying the cognitive processes underlying feigned cognitive impairment and encourage further research.

Stronger Together: The Wechsler Adult Intelligence Scale-Fourth Edition as a Multivariate Performance Validity Test in Patients with Traumatic Brain Injury

OBJECTIVE

This study was designed to evaluate the classification accuracy of a multivariate model of performance validity assessment using embedded validity indicators (EVIs) within the Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV).

METHOD

Archival data were collected from 100 adults with traumatic brain injury (TBI) consecutively referred for neuropsychological assessment in a clinical setting. The classification accuracy of previously published individual EVIs nested within the WAIS-IV and a composite measure based on six independent EVIs were evaluated against psychometrically defined non-credible performance.

RESULTS

Univariate validity cutoffs based on age-corrected scaled scores on Coding, Symbol Search, Digit Span, Letter-Number-Sequencing, Vocabulary minus Digit Span, and Coding minus Symbol Search were strong predictors of psychometrically defined non-credible responding. Failing ≥3 of these six EVIs at the liberal cutoff improved specificity (.91-.95) over univariate cutoffs (.78-.93). Conversely, failing ≥2 EVIs at the more conservative cutoff increased and stabilized sensitivity (.43-.67) compared to univariate cutoffs (.11-.63) while maintaining consistently high specificity (.93-.95).

CONCLUSIONS

In addition to being a widely used test of cognitive functioning, the WAIS-IV can also function as a measure of performance validity. Consistent with previous research, combining information from multiple EVIs enhanced the classification accuracy of individual cutoffs and provided more stable parameter estimates. If the current findings are replicated in larger, diagnostically and demographically heterogeneous samples, the WAIS-IV has the potential to become a powerful multivariate model of performance validity assessment.

BRIEF SUMMARY

Using a combination of multiple performance validity indicators embedded within the subtests of theWechsler Adult Intelligence Scale, the credibility of the response set can be establishedwith a high level of confidence. Multivariatemodels improve classification accuracy over individual tests. Relying on existing test data is a cost-effective approach to performance validity assessment.

Aggregating validity indicators: The salience of domain specificity and the indeterminate range in multivariate models of performance validity assessment

This study was designed to examine the "domain specificity" hypothesis in performance validity tests (PVTs) and the epistemological status of an "indeterminate range" when evaluating the credibility of a neuropsychological profile using a multivariate model of performance validity assessment. While previous research suggests that aggregating PVTs produces superior classification accuracy compared to individual instruments, the effect of the congruence between the criterion and predictor variable on signal detection and the issue of classifying borderline cases remain understudied. Data from a mixed clinical sample of 234 adults referred for cognitive evaluation (M_Age = 46.6; M_Education = 13.5) were collected. Two validity composites were created: one based on five verbal PVTs (EI-5_VER) and one based on five nonverbal PVTs (EI-5_NV) and compared against several other PVTs. Overall, language-based tests of cognitive ability were more sensitive to elevations on the EI-5_VER compared to visual-perceptual tests; whereas, the opposite was observed with the EI-5_NV. However, the match between predictor and criterion variable had a more complex relationship with classification accuracy, suggesting the confluence of multiple factors (sensory modality, cognitive domain, testing paradigm). An "indeterminate range" of performance validity emerged that was distinctly different from both the Pass and the Fail group. Trichotomized criterion PVTs (Pass-Borderline-Fail) had a negative linear relationship with performance on tests of cognitive ability, providing further support for an "in-between" category separating the unequivocal Pass and unequivocal Fail classification range. The choice of criterion variable can influence classification accuracy in PVT research. Establishing a Borderline range between Pass and Fail more accurately reflected the distribution of scores on multiple PVTs. The traditional binary classification system imposes an artificial dichotomy on PVTs that was not fully supported by the data. Accepting "indeterminate" as a legitimate third outcome of performance validity assessment has the potential to improve the clinical utility of PVTs and defuse debates regarding "near-Passes" and "soft Fails."

Performance Invalidity Base Rates Among Healthy Undergraduate Research Participants

Few studies have examined base rates of suboptimal effort among healthy, undergraduate students recruited for neuropsychological research. An and colleagues (2012, Conducting research with non-clinical healthy undergraduates: Does effort play a role in neuropsychological test performance? Archives of Clinical Neuropsychology, 27, 849-857) reported high rates of performance invalidity (30.8%-55.6%), calling into question the validity of findings generated from samples of college students. In contrast, subsequent studies have reported much lower base rates ranging from 2.6% to 12%. The present study replicated and extended previous work by examining the performance of 108 healthy undergraduates on the Dot Counting Test, Victoria Symptom Validity Test, Word Memory Test, and a brief battery of neuropsychological measures. During initial testing, 8.3% of the sample scored below cutoffs on at least one Performance Validity Test, while 3.7% were classified as invalid at Time 2 (M interval = 34.4 days). The present findings add to a growing number of studies that suggest performance invalidity base rates in samples of non-clinical, healthy college students are much lower than An and colleagues initial findings. Although suboptimal effort is much less problematic than suggested by An and colleagues, recent reports as high as 12% indicate including measures of effort may be of value when using college students as participants. Methodological issues and recommendations for future research are presented.