Factor analytic studies of the neo personality inventory and the five-factor model: The problem of high structural complexity and conceptual indeterminacy

Abstract

The purpose of the study was threefold: (1) to analyze the NEO Personality Inventory (NEO-PI) and the five-factor model—using both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA)—with the aim of establishing a Norwegian version of the instrument; (2) to compare the results of the analyses with the original NEO-PI simple structure five-factor models as well as later revisions of the model; and (3) to discuss causes of variations that are found in the NEO-PI literature on structural representations of personality traits. The sample comprised 973 subjects representative of the non-institutionalized Norwegian adult population. On the whole, compared with the original five-factor model and later (minor) revisions, both EFA- and CFA-strategies indicated a much more complex and fine-grained structure of items and scales. However, most of the discrepancies relative to the original model(s) were comparable to findings from recent American and German studies. Moreover, content analyses of the revised five domains as well as facets (sub-scales) indicated acceptable conceptual interpretability and plausibility. Thus, judged by conventional (mostly EFA-based) methods and criteria, the Norwegian NEO-PI version appears satisfactory in several respects. Nevertheless, it is claimed that the whole field of research is characterized by a profound conceptual indeterminacy. This problem stems largely from (1) the lack of explicit criteria-based definitions of essential concepts, and the corresponding reliance on rather arbitrary indicators (symptoms) of the alleged underlying dimensions; (2) the semantical-logical complexity of items and scales and the high probability of conceptual cross-linking; (3) the widespread tendency to define constructs on the basis of (empirical) covariance structures instead of inherent conceptual features; and (4) the masking of structural complexity by the extensive use of EFA-based methods and criteria to approximate simple structure.