topic Re: GLIMMIX and pseudo R-squares in Statistical Procedures

GLIMMIX and pseudo R-squares

deleted_user — Thu, 18 Nov 2010 09:06:46 GMT

Hi all,

Is there any way to calculate the pseudo R-squares with a GLIMMIX procedure?

Regards,

Re: GLIMMIX and pseudo R-squares

Paige — Thu, 18 Nov 2010 13:20:09 GMT

The only way I can get numbers like R-squared is to output the residual values to a data set and then calculate R-squared from those.

Re: GLIMMIX and pseudo R-squares

lvm — Fri, 19 Nov 2010 14:34:03 GMT

When there are multiple random-effect terms in a model, the simple concept of the R^2 breaks down. That is, there is no natural or unambiguous concept of the usual R^2 with two or more variance-covariance terms in a model. Various authors have proposed R^2-TYPE statistics, based on either the marginal or conditional residuals, or based on differences of log-likelihoods. I suggest you read Liu et al. 2008. J. Appl. Stat. 35:1081-1092, or Vonesh et al. 1996. Biometrics 52:572-587. One has to calculate these brute-force with SAS, by first outputting the residuals into a data file.

Re: GLIMMIX and pseudo R-squares

deleted_user — Mon, 22 Nov 2010 12:11:39 GMT

Thank you LVM and paige !