Page 1 of 1

Estimates multivariate regression analysis using RIGLS

Posted: Fri May 22, 2020 5:26 pm
by ArthurJefffrey
Hi All,
I try to run the next multivariate regression analysis using RIGLS, but as you see the estimates and SE's are the same for all. Does anyone know why this is the same (despite the fact that different variables are included)? I also see that the cases in uses differ (352 of 369 cases) versus (120 of 123 cases)
Thanks

Re: Estimates multivariate regression analysis using RIGLS

Posted: Tue Jun 16, 2020 2:12 pm
by ChrisCharlton
The (352 of 369 cases in uses) indicates that 17 rows of the expanded data contain missing values, whereas the (120 of 123) cases indicates that 3 of the participants are entirely missing. Having the same coefficients and SE's could be a coincidence, you could increase the display precision via the options->numbers menu to check that they are really the same. Have you checked the data to see whether ademN1 actually has different values between the responses?