I've run xtlogit and xtprobit regressions that yield very similar results. Would it be fine if I use the Wald Chi2 to justify why I chose one over the other? In this case, I chose the xtprobit results since its Wald Chi2 results were higher.