|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
Interpret coefficients for regression... |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
|
When the intercept is set at the final time point and it is regressed on several variables, how does the meaning of those coefficients differ from a regression on the observed value at that time point? i s | v1@-2 v2@-1 v3@0; s on x3; i on x1 x2; vs i s | v1@-2 v2@-1 v3@0; s on x3; v3 on x1 x2; |
|
|
i = b1x1 + b2x2 +e v3 = i + e3 The difference between i and v3 is e3. Using v3 ON x1 x2 says that i is not directly related to x1 and x2 which may be a model misspecification. |
|
Back to top |
|
|