RESUMO
Personal growth initiative has been shown to be an important predictor of psychological health. It is currently measured by the Personal Growth Initiative Scale-II (PGIS-II), which consists of 4 interrelated factors. Past research across various samples has consistently selected the 4-factor model as the best fit for the data compared to single-factor and second-order models. However, its fit has typically been adequate (not strong), and, to date, no research has examined alternate factor structures, such as a bifactor solution. The current study examined 4 theoretically informed potential models-single-factor, 4-factor, second-order 4 factor, and bifactor-across 3 samples drawn from different populations: 223 college students, 307 Mechanical Turk participants, and 281 clinical therapy clients. Across all 3 samples, the bifactor model was the best fit for the data, and tests of multigroup invariance indicated this model was invariant through the scalar level. Finally, analyses of the explained common variance and percentage of uncontaminated correlations indicated that the PGIS-II can be appropriately modeled unidimensionally. (PsycINFO Database Record