When you design an analysis, you postulate certain effects (eg, diagnosis, age). Each of these effects gets a column in the design matrix (eg, normal, AD, age or young/old). Each column has a regression coefficient that gets fit in the analysis that indicates the size of each of the effects. To test a hypothesis, you need to state the hypothesis in terms of the regression coefficients of the effects, eg, AD != normal, which can be translated into AD-normal ?= 0. The values in the contrast represent weights in the above equation (eg, weight for AD = 1, weight for normal= -1).

doug

On 02/24/2014 10:16 AM, Jonathan Holt wrote:

Hi all,

I humbly ask for an explanation of contrast vectors, specifically what do 0, 1, -1 indicate, exactly?

best, Jon _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer