Search

Research on personality traits predicting life outcomes has typically been investigated using the Big Five factors and only occasionally their facets. However, recent research suggests that the use of items (reflecting personality nuances) can account for more predictive variance. The aim of the present study was to examine the predictive validity for various life outcomes comparing the hierarchical levels (factors, facets, and nuances) of the personality trait structure. These were measured using one of the publicly available instruments, IPIP-NEO-120, in a Swedish sample (N = 440). Confirmatory Factor Analyses (CFA) were performed to confirm the structures of the Big Five levels, and we used Elastic Net Regressions (ENR; with 10-fold cross-validation and shrinkage parameter), trained and applied for prediction in two separate samples. The results showed that nuances (item-level models) on average provided greater explained variance (34%) than facets (22.5%) and factors (12%) for all six life outcome predictions. Findings suggest that there may be psychometric value to using the lowest item-level of personality trait measurements. Implications, limitations, and directions for future research are discussed. [ABSTRACT FROM AUTHOR]