Faking Effects on the Factor Structure of a Quasi-Ipsative Forced-Choice Personality Inventory

• Published in: Revista de psicología del trabajo y de las organizaciones Vol. 37; no. 1; pp. 1 - 10
• Main Authors: Martínez, Alexandra, Moscoso, Silvia, Lado, Mario
• Format: Journal Article
• Language: English
• Published: Madrid Revista de Psicología del Trabajo y de las Organizaciones 01.04.2021; Colegio Oficial de Psicólogos de Madrid
• Subjects: construct validity; faking; five-factor model; inventories; Job performance; measurement equivalence/ invariance; Meta-analysis; Occupational psychology; Personality; Personality tests; Quantitative psychology; quasi-ipsative forced-choice; Questionnaires; Validation studies; Validity
• ISSN: 1576-5962; 2174-0534
• DOI: 10.5093/jwop2021a7

Research has shown that faking behavior affects the factor structure of single-stimulus (SS) personality measures. However, no published research has analyzed the effects of this phenomenon on the factor structure of forced-choice (FC) personality inventories. This study examines the effects of faking, induced in a laboratory setting, on the construct validity of a quasi-ipsative FC personality inventory based on the Five-Factor Model. It also examines the moderator effect of the type of experimental design (between-subject and within-subject design) on factor analyses. The results showed that (a) data fit to a structure of five-factors in the two conditions (honest and faking) in both experimental designs; (b) model fit indices are also good or excellent in all cases; and (c) Burt-Tucker’s congruence coefficients between convergent factors of conditions analyzed are very high. These findings provide evidence that the quasi-ipsative FC format is a robust instrument that controls the effects of faking on factor structure. Finally, we discuss theoretical and practical implications of these findings for personnel selection and assessment.