Hello Hooman,

Note that, since your DV (gateway) is categorical/dichotomous, a logistic regression solution was applied. In LR, there is no direct equivalent to variance accounted for (R-squared). In LR, the log-likelihood value of the fitted model vs. that of the null model (no predictors) is the indicator of explanatory power (closer to zero is better, all other things equal).

However, the output indicates three "pseudo-R-squared" values (McFadden, Cox-Snell, and Nagelkirk). These are differently defined (as is obvious from the disagreement in values), but are intended to give an impression of the degree of explanatory power of the model. For me, the McFadden estimate behaves better than the others, but, to each their own.

Good luck with your work.

Good point David