Hello all.
For a project, I was attempting to run a logit regression generating phats for the probability of being female based on various characteristics. I wanted to construct a common support along the male and female divide but when I attempted to run a logit regression I received "No Observations." In my data, the majority of variables are discrete levels instead of being continuous. For example, education is measured by primary, secondary, illiterate, and post grad instead of the years of schooling completed. From what I can tell, whenever I only run the logit with continuous independent variables Stata gives me the regression's results. Is there an additional way I can conduct a regression with variables such as education and region to estimate the probability of being female in my sample?
For a project, I was attempting to run a logit regression generating phats for the probability of being female based on various characteristics. I wanted to construct a common support along the male and female divide but when I attempted to run a logit regression I received "No Observations." In my data, the majority of variables are discrete levels instead of being continuous. For example, education is measured by primary, secondary, illiterate, and post grad instead of the years of schooling completed. From what I can tell, whenever I only run the logit with continuous independent variables Stata gives me the regression's results. Is there an additional way I can conduct a regression with variables such as education and region to estimate the probability of being female in my sample?
Comment