This is a great video! Glad to see such new content on the channel. Question: how do I check for multicollinearity in multiple logistic regression? I have tried the vif command and it doesn't work. Am I missing something? Can I use the ## interaction function as a de facto way of looking for multicollinearity? Thanks!
We do not currently have a postestimation command to test for collinearity after *mlogit* . We are not aware of any community-contributed command either. Notice that before performing any regression estimation, Stata uses the command *_rmcoll* to check for collinearity. This command identifies and provides the group of non-collinear variables from a list. You could check for collinearity yourself by using *_rmcoll* (for details on the syntax see the entry at www.stata.com/manuals/p_rmcoll.pdf ). You may determine the group of collinear regressors ( with *_rmcoll* ), and then drop them from your specification. If you have any further questions, contact us at tech-support@stata.com and we will be happy to help out.
how to use sequential logit model. make a video of this asl well. my sequential logit model involves three stages and i do not know how to use it on stata. please help me out
PLEASE provide tutorials on non linear models, non parametric with multiple predictors. It is always logistic and binomial and there are already many tutorial videos.