[VoxBo] significance testing

[Daniel Y Kimberg]

Ling Wong wrote:

> I've got a conceptual question.  We are looking at beta values for a
> cue effect in a group (collapsed across time), and want to see if
> those beta values (through several different ROIs) are significantly
> different from zero.  For example, we expect a certain type of cue
> effect to be significant in some ROIs but not others.  What is the
> best way to test this?

I'm not sure I understand this perfectly, but it sounds like you
should be able to get by with a straightforward t-test.  If you have
effect size estimates for each subject in a set of ROIs, you can see
if that effect size is significantly different from zero with a
one-sample t-test.  If your hypothesis concerns a difference between
regions, then you would probably want the two-sample test for unequal
variances, although this kind of comparison often raises more
questions than it answers.

dan