Department of Psychology, University of Amsterdam.
Psychol Assess. 2017 Sep;29(9):1111-1119. doi: 10.1037/pas0000408. Epub 2016 Oct 31.
A growing body of research suggests that the same general dimensions can describe normal and pathological personality, but most of the supporting evidence is exploratory. We aim to determine in a confirmatory framework the extent to which responses on the Multidimensional Personality Questionnaire (MPQ) are identical across general and clinical samples. We tested the Dutch brief form of the MPQ (MPQ-BF-NL) for measurement invariance across a general population subsample (N = 365) and a clinical sample (N = 365), using Multiple Group Confirmatory Factor Analysis (MGCFA) and Multiple Group Exploratory Structural Equation Modeling (MGESEM). As an omnibus personality test, the MPQ-BF-NL revealed strict invariance, indicating absence of bias. Unidimensional per scale tests for measurement invariance revealed that 10% of items appeared to contain bias across samples. Item bias only affected the scale interpretation of Achievement, with individuals from the clinical sample more readily admitting to put high demands on themselves than individuals from the general sample, regardless of trait level. This formal test of equivalence provides strong evidence for the common structure of normal and pathological personality and lends further support to the clinical utility of the MPQ. (PsycINFO Database Record
越来越多的研究表明，相同的一般维度可以描述正常和病态的人格，但大多数支持证据都是探索性的。我们的目的是在一个验证性的框架内确定多维度人格问卷（MPQ）的反应在一般样本和临床样本中是否相同。我们使用多组验证性因子分析（MGCFA）和多组探索性结构方程建模（MGESEM），对多维度人格问卷的荷兰简短形式（MPQ-BF-NL）在一般人群子样本（N=365）和临床样本（N=365）中的测量不变性进行了测试。作为一种全面的人格测试，MPQ-BF-NL 显示出严格的不变性，表明不存在偏差。单尺度的测量不变性检验表明，10%的项目在样本之间似乎存在偏差。项目偏差仅影响成就量表的解释，与一般样本相比，临床样本中的个体更容易承认对自己的要求很高，而不管特质水平如何。这种等价性的正式检验为正常和病态人格的共同结构提供了强有力的证据，并进一步支持了 MPQ 的临床效用。
