# STATA- dummy variables

#### belkisa

##### New Member
Hey all!

I have a slight problem. I have introduced 3 dummy variables, which are basically describing the same thing. In total, I have 4 groups but in order to avoid dummy variable trap, I excluded one group and left with three.
Now when i run regression (different kinds: OLS, GLS, FE, RE), stata gives me coefficients and corresponding p values for these variables but I have problems interpreting them.

Maybe I should just run three separate regressions with the conditions for each dummy? But then my panel gets quite short...

Any ideas?

#### stekun

##### New Member
Not sure I entirely follow your question. Typically, the way to interpret a dummy coefficient is to think of it as the difference in intercept relative to the omitted group. (i.e. if you have two dummies for male and female, and omit male, and regress, say, income on these, the constant term will tell you the average income for males and the constant + coefficient will tell you the average income for females.)

If that's not helpful, perhaps you can provide a little more information about your regressions and exactly what aspect of your results are hard to interpret, and that may help me understand the problem better.