In this article, a proposed Bayesian extension of the generalized beta spatial regression models is applied to the analysis of the quality of education in Colombia. We briefly revise the beta distribution and describe the joint modeling approach for the mean and dispersion parameters in the spatial regression models' setting. Finally, we motivate the need for the innovative use of the generalized beta spatial regression model to the study of the performance of school children in Mathematics and Language, as well as in the analysis of illiteracy data, and present its results in the context of evaluating the quality of education in Colombia.