mean of the residuals

[lisa carnicelli]

I have a question. I am dealing with monthly electricity production. After detrending my variables by keeping the residuals I calculate the mean for the two countries, and I get the same result but with an opposite sign. For example, I would get 18 for country A and -18 for country B. is that normal or am I doing something wrong ?

[Nick Cox]

Assuming then equal number of observations for each country, your residuals average zero. Without knowing how you detrended, one can't be certain, but that sounds expected.

[lisa carnicelli]

I am working with monthly panel data and my model suffers from seasonality. So I have created a dummy for each month, and I have regressed the variable on the monthly dummies. Finally I saved the residuals.

this is the command:

regress total Aug Sep Oct Nov Dec Jan Feb Mar May Apr Jun, r
predict totG, resid

[Nick Cox]

Mean zero for residuals is guaranteed in principle as a consequence of what you are doing. That should be explained somewhere in most if not all regression texts.