The Big Five were originally conceived as orthogonal dimensions and the top of the personality trait hierarchy, but they have been found to be regularly correlated, and two higher-order factors or metatraits have been identified above them, often labeled Stability (conscientiousness, agreeableness, and low neuroticism) and Plasticity (extraversion and openness/intellect). Skeptics have argued that the metatraits are likely to be method artifacts rather than substantive trait dimensions. In meta-analytic data and in American and Croatian samples, using both lexical data and purpose-built measures of the Big Five, we investigated whether correlations among the Big Five remain after controlling for three sources of artifactual correlation among variables: (a) acquiescence bias, which we controlled through ipsatization; (b) evaluative consistency bias (social desirability) and other rater biases, which we controlled through the use of multiple raters; and (c) blended variables, which we controlled through the use of exploratory factor models that allow cross-loadings, thereby allowing blended variance in a given item or scale to be accounted for in the factor loadings rather than in the correlations among factors. Even after controlling for all sources of artifact simultaneously, the Big Five remained correlated at the latent level, and the correlations revealed the standard metatraits. The metatraits were not consistently correlated, providing some evidence against a general factor of personality. Our results suggest that the Big Five are genuinely correlated and that Stability and Plasticity are real trait tendencies. Hence, research aimed at understanding correlates, causes, and consequences of the metatraits remains important for psychology. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Watson et al. (Mon,) studied this question.