- CFA Exams
- Forums
- General Forum
- Topic: Adjusted R^2
| Author | Topic: Adjusted R^2 |
|---|---|
| yfmoan @2005-04-01 18:24:17 |
Is it always true that R^2adj<1? Is it always true that R^2adj>0? Give reasons |
| busquelo @2005-04-06 09:44:31 |
You got it . Adjusted always smaller than non-adjusted when doing multiple regression analysis. Mathematically, 1-[(n-1)/(n-k-1)*(1-R^2)] so as you pile on independent variables, you are multiplying the quantity (1-R^2) by a bigger ratio, so subtracting that from one becomes a smaller number, it has to. |
| yfmoan @2005-04-07 08:54:46 |
isn't it possible for adjusted r2 to be greater than unadjusted r2? I think the deal is that the vast majority of the time it will be lower but not always. You'd better check that. CAPM or Ted can correct me if I'm wrong about that I'm sure. |
