> Say I am contrasting two conditions, is the p-value based on the estimated beta of the
> two condition and the estimated standard deviation of the betas?

Right.

> (1) condition A is bigger than condition B AND condition A is bigger than condition C

Do two separate tests:

A vs B
A vs C

Then the intersection of A > B and A > C would be what you're looking for.

> (2) condition A is bigger than condition B OR condition A is bigger than condition C

Use two lines of weights for A, B, and C:

1 -1 0
1 0 -1

Combining the F-stat and the two separate t-tests should give you what you want.

> (3) condition A is bigger than the average of condition B and condition C

Set their weights with 1, 0.5, and 0.5.

> I know that you can test if different levels of one factor give different response by
> doing ANOVA on group level. Is it possible to do this on individual level? t-test
> can probably give answers to the above questions to test, but if I just want to
> know whether there is any modulation by the factor, how to do it?

With three levels (e.g., A, B, and C) for the factor you can set up the testing matrix with gltSym with two lines of weights

1 -1 0
1 0 -1

Of course the weighting is not unique, and you can assign them in different ways. This can be generalized to any number of levels.

Gang



Edited 1 time(s). Last edit at 12/18/2013 04:37PM by Gang.