One advantage of “classical’ logistic regression and especially the odds ratio for the independent variables is that you can explain to students and colleagues very graphic and vivid that if – on the independent side – onelevel of the variable changes to the next level (e.g. female to male) the probability to become addicted (e.g., the dependent dichotomous variable) is 3.5 times higher (or lower). Is this statement at all possible to make with outputs of your great (and I mean that) tool?
Yes, you’re right regarding the interpretability of the classical logistic regression model, especially if everybody understands the same concepts. With the logistic regression models created by GeneXproTools you won’t be able to explain the importance of each independent variable using that framework, but you can use other statistics and analyses like the correlation analysis and the variable importance chart.