I am conducting a regression discontinuity design where the dependent variable is continuous while control variables are categorical (except the treatment variable). In the usual regression analysis, categorical variables are put as i.cat_variable in STATA. In regression discontinuity analysis using STATA, should I put i.cat_variable or just put cat_variable?
My bet is that not putting the i. there will treat the variable in question as numerical, so i.cat would be correct :-)
The i. prefix treats the variable as an indicator (so the categories are separate and not ordered) while c. forces it to be treated as continuous.
So if there is a trend across the education categories, treating it as an indictor will miss this. Treating it as continuous will give you the average effect of a 1-category increase in educational level.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression