TYPES OF PERSONALITY BASED ON THE BIG FIVE MODEL: COMPARISON OF SOLUTIONS OBTAINED BY CLUSTER ANALYSIS AND LATENT PROFILE ANALYSIS

Abstract

The aim of this research was to determine the number and content of personality types based on participants’ scores on the dimensions of the Big Five model. Both cluster analysis and latent profile analysis were used, so the second goal was to compare the solutions obtained by these two methods. Sample consisted of 842 participants of both sexes, aged 18-68. They completed the BFI which includes 44 items with a five-point Likert type scale. Three clusters were extracted by means of a double cross-validation procedure. Latent profile analysis also extracted three classes as optimal. In both cases, the types could be named restrained, resilient, and unadapted. The restrained have average scores on all dimensions, the resilient have low scores on Neuroticism, and high on all other dimensions, while the unadapted participants have high scores on Neuroticism and low scores on all other dimension. Value of Cohen’s kappa coefficient (κ = .70) indicates that the congruence between these two solutions is high. The resilient profile, usually extracted within the dominant ARC typology, was found in this study as well, which indicates its cross-cultural stability. However, the restrained and the unadapted profile were not completely compatible with the hypercontrolled and hypocontrolled prototypes from the ARC typology, although the unadapted profile, often named non-desirable, was extracted in previous research. The results indicate a significant, although not complete, correspondance of solution contents extracted by the two used methods. Also, certain differences between personality types extracted in a Serbian sample and the solutions identified in other countries were noted.