Given the dichotomous nature of the dependent variable, a logistic regression model is proposed, which is estimated using aprobit model, using the Stata14 statistical package. The coefficients of the variables obtained in the probit estimation should be interpreted in probabilistic and marginal terms, that is, as the variation of the probability of the dependent variable before a variation of each explanatory variable, keeping the rest of the variables constant. Therefore, instead of beta coefficients (linear regression), marginal effects are used.