I want to run a linear probability model with cross-sectional data. Data was gathered from 100.000 individuals from 100 different countries. I want to examine the impact of various categorical variables on a binary outcome variable. Primarily I am interested in the effect of gender (and respectively interaction effects of gender with other predictors) on the dependent variable. Is it necessary to include something like country fixed effects? If yes, is it sufficient to include a dummy variable for country into my regression? With 100 different countries included the output would be kind of overwhelming. Is it possible to somehow simplify the results?
2.1m questions
2.1m answers
60 comments
57.0k users