Dear Juraj, the t-contrast [1 1 -2] will show voxels, in which the activation in condition A and (!) B was larger (i.e. one-tailed or directional) than in C. So this is something like (A+B)>C. If you want to compare each single condition of A and B with C you have to specify multiple t-contrasts (with according increase in multiple comparisons and need for Bonferroni-correction): [1 0 -1] for testing A>C [0 1 -1] for testing B>C If your aim is to compare A and B with condition C (in the sense A vs. C and B vs. C and in any direction) you should use an F-contrast, like [1 0 -1; 0 1 -1] for the hypothesis A not equal C, B not equal C. Good luck, Thilo On Thursday 16 June 2005 14:39, Juraj Kukolja wrote: > Dear statistics experts: > > Basic question: > If I have conditions A and B and want to compare both with condition C > (design matrix: [A B C]; each condition with equal trial number) in an fMRI > experiment, a common procedure would be to use contrast [1 1 -2]. > What does the "double weighting" of condition C do to it? Does this > contrast merely multiply beta values by -2, or does it multiply the trial > number of condition C by -2 and thus artificially reduce the variance? Is > this a problem for the statistical inference? > Is there any literature specifically dealing with this topic? > > Thanks in advance > > > Juraj -- Thilo Kellermann RWTH Aachen University Pauwelstr. 30 52074 Aachen Tel.: +49 (0)241 / 8089977 Fax.: +49 (0)241 / 8082401 E-Mail: [log in to unmask]