17 Performance Validity and Ethnicity

Abstract

Objective:

This study examined false positive rates of performance validity test (PVT) failure in a group of monolingual (English-speaking) White non-Hispanic/Latinx (non-HL), monolingual (English-speaking) Hispanic/Latinx (HL), and bilingual (English- and Spanish-speaking) HL patients evaluated at an academic medical center. Research on classification accuracy of embedded performance validity tests (PVTs) is limited in HL and bilingual populations. Cultural test biases or language differences could inaccurately cause scores below PVT cutoff levels.

Participants and Methods:

The project involved secondary analysis of a deidentified dataset (N=391). Participants were included if they were between the ages of 18 and 64, had data from the initial visit, had an IQ greater than or equal to 70, were not diagnosed with dementia or major or mild cognitive impairment, and identified as either White non-HL or HL. Participants were required to have completed at least two PVTs. Participants who were not administered the Test of Memory Malingering (TOMM; n=95) or who scored below a highly sensitive Trial 1 cutoff (<46; n=86) were excluded. The final sample included 210 participants, which included monolingual non-HL participants (n=114), monolingual HL participants (n=44), and bilingual HL participants (n=52). Failure rates on eight PVTs were examined by participant group: Reliable Digit Span (RDS), Auditory Verbal Learning Test Recognition (AVLT), Logical Memory Recognition (LM), Visual Reproduction Recognition (VR), Trail Making Test Ratio (TMT rat), Rey Complex Figure Test Recognition (RCFT), Semantic Word Generation (animals; SWG), and Finger Tapping (TAP).

Results:

Groups were not significantly different in age. Monolingual non-HL participants had completed more years of education than monolingual and bilingual HL groups (13.7, 12.7, and 12.8 years respectively). In the whole sample, 8.6% (n=18) failed two or more PVTs. In the monolingual non-HL group, 8.8% (n=10) failed two or more PVTs, while 9.1% (n=4) of the monolingual HL group and 7.7% (n=4) of the bilingual HL group failed two or more PVTs (n.s.). Within the monolingual non-HL test set, failure rates were above 10% on SWG (12.73%) and TAP (17.7%). Failure rates above 10% in the monolingual HL set were found on SWG (11.6%). Failure rates above 10% in bilingual HL measures were observed on SWG (13.5%) and TAP (10.8%).

Conclusions:

Total PVT failure rates did not significantly differ between groups. Across groups, performance was above a common false positive threshold of 10% on SWG. Monolingual non-HL and bilingual participants also had elevated failure rates on TAP.