Are personality measures valid for different populations? A systematic review of measurement invariance across cultures, gender, and age

Abstract

This article aims to systematically review and synthesize the existing literature that has tested at least one of three types of measurement invariance (MI) of a personality measure: cross-cultural/ethnic invariance, gender invariance, and age-group invariance. A literature search was conducted using PsycINFO, PsycARTICLES, Psychology and Behavioral Sciences Collection, and PsycTESTS databases. The initial search yielded a total of 1,647 articles, and 95 studies derived from 75 peer-reviewed articles met all inclusion criteria. None of these studies achieved scalar or strict measurement invariance across cultural/ethnic groups, but many studies established scalar or strict invariance across gender (N = 13, 44.83%) or age (N = 10, 58.82%) groups. To further investigate patterns in this literature, an ordinal logistic regression revealed that age-group and gender invariance studies achieved significantly higher levels of invariance than cross-cultural MI studies, and confirmed that studies testing invariance across larger number of groups had significantly lower MI levels. Consequences and possible reasons for a lack of invariance, and suggestions for improving practices of investigating measurement invariance are discussed.