The Practical Effects of Measurement Invariance: Gender Invariance in Two Big Five Personality Measures. (June 2020)

• Record Type: Journal Article
• Authors: Ock, Jisoo; McAbee, Samuel T.; Mulfinger, Evan; Oswald, Frederick L.
• Is Part Of: Assessment. Volume 27:Number 4(2020)
• Page Start: 657
• Page End: 674
• Publication Date: 2020-06
• Subjects: measurement invariance -- Big Five -- model fit index -- gender differences -- practical significance; Psychological tests -- Periodicals; Psychodiagnostics -- Periodicals; Personality tests -- Periodicals; 150.287
• DOI: 10.1177/1073191119885018
• Languages: English
• ISSNs: 1073-1911

Abstract

The essence of measurement invariance (MI) analysis is to test the assumption that observed scores on a scale accurately reflect respondents' standings on a measured construct. Based on exploratory structural equation modeling, the current study examines gender-based MI in two Big Five measures of personality: the Mini-IPIP, and the Big Five Inventory (BFI) facet scales. We report results for MI based on both model fit indices and a practical significance index that quantifies the extent of noninvariance (i.e., dMACS ). From the latter, we partition the observed group mean differences in scale scores into construct-irrelevant group differences versus construct-relevant group differences. In measures of Agreeableness, Conscientiousness, Extraversion, and Neuroticism across instruments, results supported metric invariance but not scalar invariance. That said, findings of statistical noninvariance were generally small in terms of practical effects, although some notable variability in the effects was evident. Overall, the current results provide evidence regarding gender-based MI of Big Five personality measures that are more detailed than that provided in past work. More generally, this study also provides useful guidance for future researchers investigating both the statistical and practical significance of measurement invariance.