RESUMO
Background: Self-efficacy is a psychological variable closely related to healthy behaviors. One of the most widely used instruments to measure this variable is the general self-efficacy scale (GSES). In Chile, the only psychometric study of this scale has important analytical limitations. Aim: To assess reliability and validity of GSES in a large Chilean sample. Material and Methods: The GSES and a self-esteem scale were applied to 2995 participants aged 11 to 76 years (60% women). Results: The scale showed optimal levels of homogeneity and internal consistency. The exploratory factor analysis and the parallel analysis suggested a one-dimensional internal structure. However, this model showed a mediocre fit in the confirmatory factor analysis due to the residual correlation between a pair of items with high semantic similarity. A one-dimensional model specifying this residual correlation attained an acceptable fit in a new confirmatory factor analysis. The new model reached strict invariance according to sex, and partial strict invariance according to age. The latent factor of the GSES showed a positive association with the latent factor of a self-esteem scale. Conclusions: The GSES is an adequate scale to measure general self-efficacy in the Chilean population and allows unbiased comparisons based on sex or age.