Confirmatory factor analysis of the M5-50: An implementation of the International Personality Item Pool item set.

Abstract

Goldberg's International Personality Item Pool (IPIP; Goldberg, 1999) provides researchers with publicdomain, free-access personality measurement scales that are proxies of well-established published scales. One of the more commonly used IPIP sets employs 50 items to measure the 5 broad domains of the 5-factor model, with 10 items per factor. The M5-50 (McCord, 2002) is a specific ordering and presentation of this 50-item set. Using data from a sample of 760 faculty, staff, and students at a midsized university, the authors assessed the reliability and construct validity of the M5-50. Cronbach's alphas on the 5 scales ranged from acceptable to excellent. Confirmatory factor analysis indicated reasonably good model fit. Researchers who wish to measure personality would be well advised to consider using the M5-50.

Taking Personality Seriously

“…There are 10 items per factor, 5 of which are reverse-coded to ensure that participants do not provide the same answer to each question. This publically available personality test is a valid and reliable indicator of the FFM of personality (Socha, Cooper, & McCord, 2010). The validity of self-report personality instruments such as these has been well established, both in terms of rational and empirical support.…”

Development and Initial Validation of the Five-Factor Model Adolescent Personality Questionnaire (FFM–APQ)

“…While some researchers have found reasonable FFM model fits using CFA (Socha, Cooper, & McCord, 2010), due to inconclusive or poorly-fitting models, others have queried the utility of CFA for deriving the FFM dimensions (Borkenau & Ostendorf, 1990;Donnellan et al, 2006;McCrae, Zonderman, Costa, Bond, & Paunonen, 1996;Parker, Bagby, & Summerfeldt, 1993;Vassend & Skrondal, 1997). Even EFA-developed scales may not be supported by CFA (Marsh et al, 2009(Marsh et al, , 2010McCrae et al, 1996).…”

The M5–PS–35: A Five-Factor Personality Questionnaire for Preschool Children

“…The M5-PS-35's results were also similar to the adult M5-50 (Socha, Cooper, & McCord, 2010) and the 60-item NEO-FFI (Marsh et al, 2010). Each item loaded significantly on its factor for the M5-PS-35 (see Table 3 for CFA structure coefficients on the target factors; coefficients on the nontarget factors were set to zero).…”

“…First, we should stress the fact that CFA was used as a descriptive comparative technique, helping to improve the instrument, rather than as a pure test of model fit. Although there are some instances in which a five-factor instrument does indeed meet, or approach, ideal CFA statistics (e.g., Socha et al, 2010), in most cases the recommended cutoffs for theory testing are not met when testing five-factor and other personality measurement instruments (Marsh et al, 2010). Marsh et al (2010) addressed this issue very specifically, noting that exploratory factor analysis was used extensively in the development of the FFT, but CFA has failed to provide clear support for the model.…”