Testing the generality of the general factor of personality: An exploratory bifactor approach

Abstract Objective To investigate whether the psychometric properties of the general factor of personality (GFP) obtained through self-reported measures support its interpretation as a substantive dimension of general order. Method We estimated oblique and orthogonal bifactor exploratory structural equation models of the Big Five. Results The GFP explained considerably less variance than the five group factors and showed poor model-based reliability. The pattern of GFP loadings was consistent with those of a reverse-keyed wording factor. When related to an external variable (dispositional optimism) the GFP was primarily associated to method variance, and not to the substantive criterion. Conclusions Although there is a certain degree of variance common to most behavioral indicators of personality, its properties are not compatible with an interpretation of the GFP as a reliable and psychometrically meaningful general factor of personality.