Abstract
The general objective of this thesis is to apply modern multivariate methods in a Norwegian twin sample in order to sharpen and nuance crucial aspects of how we understand the reality and fundamental properties of the Big Five personality traits (i.e., the ontology) and their origin and development (i.e., the etiology). To achieve the general objective, our first aim was to assess the Big Five facets, which were used as building blocks in the models applied in our analysis. This appraisal was performed in two steps. The first step was to examine the phenotypic facet dimensionality in the current sample. A five-factor structure emerged in our sample. However, significant alterations of two of the proposed main domains were evident, augmenting questions about the alleged universality of the Big Five traits. The second step was to estimate the heritability and its relation to construct unity. The moderate correlation between the wide-ranging heritability and Cronbach’s alpha estimates, together with the apparent cross-loadings, raise essential ontological questions regarding the applied building blocks of the Big Five. The second aim was to contribute to the knowledge of the traits’ etiology by investigating the sources of variation underpinning the expression of personality. To address this aim, additive genetic and non-shared environmental facet correlation matrices were extracted from a Cholesky twin design model, and principal component analysis was applied to estimate factors. Five factors were extracted in the genetic correlation matrix, which greatly resembled the phenotypic rotation. Four factors emerged in the environmental matrix, which also resembled the phenotypic rotation with the exception of the fusion of two factors. The third aim targeted the ontology of the Big Five traits, through scrutiny of the fundamental interpretation and application of the five-factor model. Ever since Allport (1931) postulated that “a trait has more than nominal existence” (p. 368), the debate regarding the nature of personality traits has, to a certain extent, been buried by an implicit or explicit acceptance of Allport’s position. Although buried, the unsettled assumption has haunted the research field, and in this thesis, we dig up this nearly one-hundred-year-old statement by comparing two profound theoretical perspectives, which can be regarded as contrasting ontological positions: the realist interpretation, which considers the Big Five dimensions to be veridical entities that coincide with reality and emerge from a biological basis, ultimately anchored in genes, and the constructivist interpretation, which assumes that personality traits identified through factor analysis mainly reflect semantic clusters in the language. To assess these interpretations, common and independent pathways models were compared to test the five-factor model’s ability to mediate genetic and environmental contributions. The independent pathways model fits the data comparatively better, indicating that the five factors do not mediate genetic and environmental contributions to personality trait constructs. The model that fit the quantitative genetic data best was an alternative local etiological independent pathways model. The results reaffirmed the ambiguous universality and equivocal facets of the Big Five, which question the origin, development, structure, and nature of the traits as proposed in the five-factor theory. The etiological exploration of genetic and environmental components indicates that both endogenous and systematic exogenous influence structure the Big Five traits. Ultimately, the results from the pointed operationalization of fundamental ontological positions are unsupportive of an interpretation of the Big Five traits as causal explanations for aggregated thoughts, emotions, and behavior.