Date of Award

2-7-2019

Publication Type

Doctoral Thesis

Degree Name

Ph.D.

Department

Psychology

First Advisor

Laszlo A Erdodi

Keywords

ESL, Experimental malingering, Limited English proficiency, Neuropsychological testing, Performance validity testing

Rights

info:eu-repo/semantics/openAccess

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Abstract

Background Performance validity tests (PVTs) are an integral component of neuropsychological assessments. Despite the growing literature on PVTs, little research has focused on how these instruments perform in individuals with limited English proficiency (LEP). Indeed, the majority of PVTs have been developed and validated with individuals who are native speakers of English (NSE), and their psychometric properties have not yet been established for an LEP population. Objectives The current dissertation aimed to (1) determine the effect of LEP on PVT performance; (2) examine signal detection properties of current PVTs in individuals with LEP; and (3) develop new PVT cutoffs for this population. Methods To examine these objectives, a two-part prospective study was conducted. Part 1 consisted of using a case-control design to compare PVT performance between LEP and NSE groups. Part 2 consisted of using a single-blind, experimental-malingering design to establish classification accuracy across a battery of PVTs with an LEP sample. Participants (N = 140) were randomly assigned to either a non-malingering control or experimental-malingering condition. Research assistants, who were blinded to the experimental condition of the participant, administered a battery of neuropsychological tests containing PVTs with high verbal mediation (PVTHVM) and low verbal mediation (PVTLVM). Both a liberal cutoff, maximizing sensitivity, and a conservative cutoff, emphasizing specificity, were chosen from the literature to calculate base rates of failure (BRFail). Results Part 1. Under normal conditions (i.e., not instructed to malinger), participants with LEP had a higher BRFail on and failed more PVTHVM compared to NSE. In contrast, BRFail and number of PVTs failed were similar between groups on PVTLVM. English proficiency was highly correlated with BRFail on PVTHVM but not on PVTLVM. Part 2. Using published cutoffs, PVTLVM demonstrated good classification accuracy, while the majority of PVTHVM were not specific to malingering for the LEP sample. Adjusted cutoffs resulted in high sensitivity while maintaining adequate specificity on many PVTLVM, but an optimal balance of sensitivity and specificity was unable to be obtained on some PVTHVM regardless of how cutoffs were adjusted. Conclusions & Future Directions PVTHVM increased false-positive errors for individuals with LEP, as both experimental malingering and LEP produce an elevated BRFail on these tests. Although there were instrument-specific exceptions to the overall findings, it is generally recommended that examiners preclude the use of PVTHVM for individuals with LEP. The current study established new cutoffs on many PVTs that are both specific and sensitive for this population. As a field, neuropsychological testing with cultural and linguistic minorities have been identified as a prominent issue, and the need for further studies in individuals with LEP is evident. Future investigations should focus on validating the new LEP cutoffs with different demographic samples in clinical and forensic settings.

Share

COinS