Hi Steve Thank you very much for your reply. Do you mean that the randomise program can't perform a two-sample test, since it used the GLM? However, there is an example of a two-sample t-test in the randomise manual, so the example is incorrect? Thank you in advance for the clarificaiton Johnny Date: Thu, 12 Jan 2012 09:29:07 +0000 From: [log in to unmask] Subject: Re: [FSL] Confusion about the stats implemented in TBSS To: [log in to unmask] Hi - no, if you run a GLM with randomise you get a GLM, using the model and contrasts you specified - so this is not explicitly an ANOVA via residual variances. Cheers. On 12 Jan 2012, at 00:54, JohnnyZhuang wrote: Dear FSL experts, I am running a TBSS analysis to test the group differences in FA between two groups. As suggested by the TBSS manual, i performed the 2-sample t-test implemented in the randomise program to evaluate the differences in FA between patients and controls, when controlling for age and sex. Does the TBSS still perform the t-test by comparing the mean of residuals of FA between two groups, after controlling for these covariates? In other words, does the controlling part make the stats become ANOVA rather than t-test? Any help and suggestion will be appreciated. Johnny --------------------------------------------------------------------------- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ---------------------------------------------------------------------------