Dear Joyce,
It is quite possible that you have no individually significant results for these contrasts but can have results for your differential contrasts (i.e. [1 -1] and [-1 1]). They are asking different questions (e.g. for the [1 0] contrast it is whether the mean of one group is significantly _greater_ than zero). You should always look at the *corrp* images to see what is significant after correction for multiple comparisons.
All the best,
Mark
On 12 Dec 2012, at 14:25, Joyce Yu Guo <[log in to unmask]> wrote:
> I have got a further question, why there is no significant results for [0 1 ] or [1 0] contrast?
>
|