Dear List<div><br></div><div>I&#39;m wondering about the proper way to set up an analysis that includes site as a regressor, and whether this is possible in qdec?</div><div><br></div><div>My qdec.table.dat file looks like this.</div>

<div><br></div><div><div>fsid    group   site</div><div>1       grp1      1</div><div>2       grp2      2</div><div>3       grp2      2</div><div>4       grp1      3</div></div><div>.... and so on</div><div><br></div><div>

I then have .level files for group which has 2 levels, and site which has 7 levels. </div><div><br></div><div>The type of question I want to answer for example is: Is there a difference between grp1 and grp2 regressing out the effects of site?</div>

<div><br></div><div>Is this possible with Qdec? </div><div>When I select the group variable on the design tab the analysis results shows the question of the form: Is there a difference between grp1 and grp2, but there is no mention of regressing out the site effect. Looking at the design matrix X.mat it only includes 2 columns which isn&#39;t right.</div>

<div><br></div><div>If I select both group and site on the design page I get an error.</div><div><br></div><div>Do I need to put together an FSGD file instead?</div><div><br></div><div>Thanks</div><div>Darren</div><div><br>

</div><div><div><br></div>-- <br>Darren Gitelman, MD<br>Northwestern University<br>710 N. Lake Shore Dr.<br>Abbott 11th Floor<br>Chicago, IL 60611<br>Ph: (312) 908-8614<br>Fax: (312) 908-5073<br><br><br>
</div>