The objective of this diagnostic research was to identify the perceptions of indigenous communities regarding the digital divide and government proposals for their inclusion in formative processes. For this purpose, the Indigenous Leadership School was developed at the University of Atacama in 18 months. The hierarchical segmentation, classification and regression (CART) method was applied to identify and select the determining variables of ICT management, and once the variables were selected, the associations between them were determined with the multivariate method of multiple correspondence analysis (MCA). Binary logistic regression analysis was used to evaluate the probability of presenting poor ICT management according to the variable selected as the most determinant of this characteristic, which corresponded to age, dichotomized into those under 46 and those over 45 years of age. Finally, univariate analysis of descriptive statistics of quantitative and categorical variables, centralization, dispersion, histogram and bar graph statistics were estimated. One of the strong conclusions indicates that those older than 45 years of age are about three times more likely to show poor ICT skills than those younger than 46 years of age.