I should add that I&#39;m speaking from the general standpoint of how the GLM works (so for a FEAT GLM, R GLM, SPSS GLM).  Randomise with the -D option is different as it implicitly adds the column of 1s.<div><br></div><div>
Cheers,</div><div>Jeanette<br><br><div class="gmail_quote">On Mon, Mar 18, 2013 at 12:09 PM, Jeanette Mumford <span dir="ltr">&lt;<a href="mailto:[log in to unmask]" target="_blank"><a href="/cgi-bin/wa-jisc.exe?LOGON=A3%3Dind1303%26L%3DFSL%26E%3Dquoted-printable%26P%3D2472576%26B%3D--047d7bf198948ed9fe04d83626fe%26T%3Dtext%252Fhtml%3B%2520charset%3DISO-8859-1%26pending%3D" target="_parent" >[log in to unmask]</a></a>&gt;</span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Yes, there&#39;s a huge difference, one model has 2EVs and the other has 1.  This will always indicate models are not equivalent.  This is also a different problem than what you asked originally as I thought you had a group covariate as well.  Your second model would need a column of 1s to be equivalent to the first.  <div>

<div><br></div><div>Without the column of 1&#39;s you&#39;re assuming your two group&#39;s means are perfectly centered about 0.</div><div><br></div><div>Cheers,</div><div>Jeanette<div><div class="h5"><br><div><br></div>
<div><br><br><div class="gmail_quote">
On Mon, Mar 18, 2013 at 11:11 AM, Vincent Koppelmans <span dir="ltr">&lt;<a href="mailto:[log in to unmask]" target="_blank"><a href="/cgi-bin/wa-jisc.exe?LOGON=A3%3Dind1303%26L%3DFSL%26E%3Dquoted-printable%26P%3D2472576%26B%3D--047d7bf198948ed9fe04d83626fe%26T%3Dtext%252Fhtml%3B%2520charset%3DISO-8859-1%26pending%3D" target="_parent" >[log in to unmask]</a></a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">

<div style="word-wrap:break-word">Hi Jeanette,<div><br></div><div>Thanks for your help.</div><div><br></div><div>So is there any difference in the outcomes of the following two designs?</div><div><br></div><div>A)</div><div>

Group A = 0 1</div><div>Group B = 1 0</div><div><br></div><div>Design.mat</div><div>0 1</div><div>0 1</div><div>0 1</div><div>1 0</div><div>1 0</div><div>1 0</div><div><br></div><div>Design.con</div><div>1 -1</div><div>-1 1</div>

<div><br></div><div><br></div><div>B)</div><div>Group A = -1</div><div>Group B = 1</div><div><br></div><div>Design.mat </div><div>-1</div><div>-1</div><div>-1</div><div>1</div><div>1</div><div>1</div><div><br></div><div>
Design.con</div>
<div>1</div><div>-1</div><div><br></div><div>Thanks, Vincent</div><div><br>
<br><div><div>Op 18 mrt. 2013, om 12:01 heeft Jeanette Mumford &lt;<a href="mailto:[log in to unmask]" target="_blank"><a href="/cgi-bin/wa-jisc.exe?LOGON=A3%3Dind1303%26L%3DFSL%26E%3Dquoted-printable%26P%3D2472576%26B%3D--047d7bf198948ed9fe04d83626fe%26T%3Dtext%252Fhtml%3B%2520charset%3DISO-8859-1%26pending%3D" target="_parent" >[log in to unmask]</a></a>&gt; het volgende geschreven:</div><div><div><br><blockquote type="cite">

Hi,<div><br></div><div>You&#39;re on the right track. Add a covariate that is 1 for left and -1 for right.  Assuming Ev1=group1 EV2=group2 and EV3=handedness, the </div><div>[ 0 0 1] contrast will test for when left&gt;right</div>


<div>[0 0 -1] contrast will test for when right&lt;left</div><div><br></div><div>No collinearity will occur if you limit yourself to these 3EVs and you can get at all inferences of interest.</div><div><br></div><div>Cheers,</div>


<div>Jeanette<br><br><div class="gmail_quote">On Mon, Mar 18, 2013 at 10:20 AM, Vincent <span dir="ltr">&lt;<a href="mailto:[log in to unmask]" target="_blank"><a href="/cgi-bin/wa-jisc.exe?LOGON=A3%3Dind1303%26L%3DFSL%26E%3Dquoted-printable%26P%3D2472576%26B%3D--047d7bf198948ed9fe04d83626fe%26T%3Dtext%252Fhtml%3B%2520charset%3DISO-8859-1%26pending%3D" target="_parent" >[log in to unmask]</a></a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">


Dear FSL experts,<br>
<br>
When you create a design matrix using for example &quot;design_ttest2 design 7 11&quot; you will get a 2 by 18 matrix. The two columns both represent group membership, which are both needed in order to find out where group1 &gt; group2 and where group2 &gt; group1.<br>



<br>
However, when I enter a dichotomous covariate of no interest (e.g. handedness), I do not need to create two columns. Creating one column would be sufficient to model out the effect of handedness. Is that correct?<br>
<br>
If I would be interested for which voxels left &gt; right, and when right &gt; left, then I should create two columns, just like the group variable created by design_ttest2 (and change the the design.con file accordingly). Right?<br>



<br>
Now if I consider the last scenario, where there are two columns for handedness, but that are entered as covariates of no interest,  would that result in multicolinearity? How does randomise handles this?<br>
<br>
Thanks,<br>
<br>
Vincent<br>
</blockquote></div><br></div>
</blockquote></div></div></div><br></div></div></blockquote></div><br></div></div></div></div></div>
</blockquote></div><br></div>