Measurement Invariance of Three Narcissism Questionnaires Across the United States, the United Kingdom, and Germany. (January 2021)

• Record Type: Journal Article
• Authors: Wetzel, Eunike; Lang, Felix J.; Back, Mitja D.; Vecchione, Michele; Rogoza, Radoslaw; Roberts, Brent W.
• Is Part Of: Assessment. Volume 28:Number 1(2021)
• Page Start: 29
• Page End: 43
• Publication Date: 2021-01
• Subjects: narcissism -- measurement invariance -- measurement equivalence -- cross-cultural differences -- Brief Pathological Narcissism Inventory -- Narcissistic Personality Inventory -- Narcissistic Admiration and Rivalry Questionnaire; Psychological tests -- Periodicals; Psychodiagnostics -- Periodicals; Personality tests -- Periodicals; 150.287
• DOI: 10.1177/1073191120907967
• Languages: English
• ISSNs: 1073-1911

Abstract

With a recent surge of research on narcissism, narcissism questionnaires are increasingly being translated and applied in various countries. The measurement invariance of an instrument across countries is a precondition for being able to compare scores across countries. We investigated the cross-cultural measurement invariance of three narcissism questionnaires (Brief Pathological Narcissism Inventory [B-PNI], Narcissistic Personality Inventory [NPI], and Narcissistic Admiration and Rivalry Questionnaire [NARQ]) and mean-level differences across samples from the United States ( N = 2, 464), the United Kingdom ( N = 307), and Germany ( N = 925). Overall, the B-PNI and NARQ functioned equivalently for the U.S. and U.K. participants. More violations of measurement invariance were found between Germany and the combined U.S. and U.K. samples, and for the NPI. In the B-PNI and NARQ, Americans scored higher than individuals from the United Kingdom regarding agentic aspects (self-sacrificing self-enhancement, admiration), while Germans scored lower than both Americans and U.K. individuals regarding antagonistic (entitlement rage, rivalry) and neurotic (hiding the self, contingent self-esteem) aspects. More inconsistent results were found for NPI facets. When noninvariance was present, observed means yielded biased results. Thus, the degree of measurement invariance across translated instrument versions should be considered in cross-cultural comparisons, even with culturally similar countries.