I am working with a dataset of about a total of 28000 observations for over a period of 10 years. I ran a probit model for my baseline results. Both my dependent and independent variables are binary. Now i suspect that there might be a reverse causality issue in my model. So I did a IV probit in Stata. But result turned out to be insignificant (showing there is no endogeneity concerns). Then I realized I should not be using IV probit if both the variables are binary. Now I need to solve my reverse causality issue using some other method. I need help with stata commands.