topic Re: Glimmix: Using covtest to compare the matrices for G side random effects for different treatments in Statistical Procedures https://communities.sas.com/t5/Statistical-Procedures/Glimmix-Using-covtest-to-compare-the-matrices-for-G-side-random/m-p/193076#M10266 <HTML><HEAD></HEAD><BODY><P>I think the covtest general approach is the only one you can use to get at your question.&nbsp; The other options available for G side hypotheses don't necessarily fit.</P><P></P><P>One thing you could do is look at the information criteria obtained when fitting models with and without the group=treatment option.&nbsp; This would give an idea of how much information is lost by assuming that all groups have equal variances, as opposed to all groups having different variances.&nbsp; Note that this is not a statistical test that will return a probability value.</P><P></P><P>Steve Denham</P></BODY></HTML> Fri, 10 Apr 2015 17:39:29 GMT SteveDenham 2015-04-10T17:39:29Z Glimmix: Using covtest to compare the matrices for G side random effects for different treatments https://communities.sas.com/t5/Statistical-Procedures/Glimmix-Using-covtest-to-compare-the-matrices-for-G-side-random/m-p/193075#M10265 <HTML><HEAD></HEAD><BODY><P>When using Glimmix, it is possible to test if the matrices for the R side random effects for different treatments are equal. For example, if you measure three variables per individual (id), with the individual receiving one of two treatments (for example high or low food) this would be done by including: </P><P></P><P><SPAN style="font-size: 10.0pt; font-family: Courier;">random _residual_ / type=un group=treatment subject=id;</SPAN></P><P></P><P><SPAN style="font-size: 10.0pt; font-family: Courier;">Covtest homogeneity;</SPAN></P><P></P><P>Now say that individuals belong to different families (g side random effect):</P><P></P><P><SPAN style="font-size: 10.0pt; font-family: Courier;">random variable / type=un group=treatment subject=family;</SPAN></P><P></P><P><SPAN style="font-size: 10.0pt; font-family: Courier;">random _residual_ / type=un group=treatment subject=family(id);</SPAN></P><P></P><P><SPAN style="font-size: 10.0pt; font-family: Courier;">Covtest homogeneity;</SPAN></P><P></P><P></P><P>Question 1:</P><P></P><P>In the second code, what is the covtest statement actually requesting? Will this just show if the matrices for the r side random effects are different for each treatment?</P><P></P><P>Question 2:</P><P></P><P>I would like to know specifically if the matrices for the G side random effects (family) are different for each treatment, is it possible to specify this?&nbsp; Is the only way to do this by using ‘covtest general’ and then specify elements from the ‘Covariance Parameter Estimates’ table you want to compare?</P></BODY></HTML> Fri, 10 Apr 2015 14:20:47 GMT https://communities.sas.com/t5/Statistical-Procedures/Glimmix-Using-covtest-to-compare-the-matrices-for-G-side-random/m-p/193075#M10265 DIHS 2015-04-10T14:20:47Z Re: Glimmix: Using covtest to compare the matrices for G side random effects for different treatments https://communities.sas.com/t5/Statistical-Procedures/Glimmix-Using-covtest-to-compare-the-matrices-for-G-side-random/m-p/193076#M10266 <HTML><HEAD></HEAD><BODY><P>I think the covtest general approach is the only one you can use to get at your question.&nbsp; The other options available for G side hypotheses don't necessarily fit.</P><P></P><P>One thing you could do is look at the information criteria obtained when fitting models with and without the group=treatment option.&nbsp; This would give an idea of how much information is lost by assuming that all groups have equal variances, as opposed to all groups having different variances.&nbsp; Note that this is not a statistical test that will return a probability value.</P><P></P><P>Steve Denham</P></BODY></HTML> Fri, 10 Apr 2015 17:39:29 GMT https://communities.sas.com/t5/Statistical-Procedures/Glimmix-Using-covtest-to-compare-the-matrices-for-G-side-random/m-p/193076#M10266 SteveDenham 2015-04-10T17:39:29Z