Soble Jason R, Cerny Brian M, Rhoads Tasha, DeBoer Adam B, Sharp Dillon W, Ovsiew Gabriel P, Phillips Matthew S, Pesanti Stephen D, Jennette Kyle J, Resch Zachary J
Department of Psychiatry, University of Illinois College of Medicine.
Neuropsychology. 2023 Feb;37(2):218-232. doi: 10.1037/neu0000880. Epub 2022 Nov 10.
This study examined the extent to which demographic variables (i.e., age, education, premorbid IQ, sex, ethnoracial identity, and presence/absence of external incentive) affect performance validity test (PVT) performance.
This cross-sectional study examined two distinct, diverse outpatient clinical samples at an academic medical center (AMC, = 268) and a Veterans Affairs (VA) medical center ( = 111). All patients completed a battery including five PVTs. Premorbid IQ was assessed using the Test of Premorbid Functioning (TOPF) in the AMC sample.
Multiple correlations between demographic variables and individual PVT performance were statistically significant, but accompanying effect sizes were small, except for the relationship of premorbid IQ and reliable digit span (RDS). Regressions showed demographic variables accounted for 7%-11% of the variance in individual PVT scores in the AMC sample, and 6%-26% in the VA sample, premorbid IQ driving results in the AMC sample and compensation-seeking status in the VA sample. Other demographic variables did not correlate with compensation-seeking status. Additionally, premorbid IQ was found to be significantly higher in validly performing individuals compared to those performing invalidly in the AMC sample.
Most demographic factors evaluated accounted for relatively little variance in individual PVT performance and did not significantly predict overall validity categorization. Compensation-seeking status correlated with validity classification across both groups, but offers limited diagnostic utility itself compared to objective PVT scores. Premorbid IQ within the AMC group demonstrated influence on particular PVTs (i.e., RDS) reflecting the difficulty of assessing validity within low IQ populations, particularly with PVTs more strongly correlated with IQ. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
本研究考察了人口统计学变量（即年龄、教育程度、病前智商、性别、种族身份以及是否存在外部激励）对效标效度测试（PVT）表现的影响程度。
这项横断面研究考察了一家学术医疗中心（AMC，n = 268）和一家退伍军人事务（VA）医疗中心（n = 111）的两个不同且多样的门诊临床样本。所有患者都完成了一组包含五项PVT的测试。在AMC样本中，使用病前功能测试（TOPF）评估病前智商。
人口统计学变量与个体PVT表现之间的多重相关性具有统计学意义，但伴随的效应量较小，病前智商与可靠数字广度（RDS）的关系除外。回归分析表明，在AMC样本中，人口统计学变量占个体PVT分数方差的7%-11%，在VA样本中占6%-26%，在AMC样本中病前智商主导结果，在VA样本中寻求补偿状态主导结果。其他人口统计学变量与寻求补偿状态无关。此外，在AMC样本中，有效表现的个体的病前智商显著高于无效表现的个体。
所评估的大多数人口统计学因素在个体PVT表现中占的方差相对较小，并且不能显著预测总体效度分类。寻求补偿状态与两组的效度分类相关，但与客观PVT分数相比，其本身的诊断效用有限。AMC组内的病前智商对特定的PVT（即RDS）有影响，这反映了在低智商人群中评估效度的困难，特别是对于与智商相关性更强的PVT。（PsycInfo数据库记录（c）2023美国心理学会，保留所有权利）
