 |
 |
Thank you Donald,
We have looked at the average of the two conditions at the single
subject level, but it looks like "OR" (showing sum of activation
of both conditions) rather than "AND" (which would show common
activations from both). And the "AND" is what are we looking for.
What we have tried to do now is the multiplication of cons (using
ImCalc) for print and speech at first level.
Our intuition is that this should be closer to "AND" - would you
agree?
Do you think this is correct to apply second-level t-test on to
the first level products of cons? (it is doable, but we are not
sure if we are allowed to do a t-test).
We will be grateful for any comments and suggestions.
Marek
[log in to unmask]" type="cite">Marek,
What the reviewer is asking for is not possible because the conjunction does not produce a con_ image. Without the con_ image, you can't do the second-level analysis.
I think what the reviewer is asking for is to create a contrast of the average of the two conditions and to test if the average of the two conditions is different between the groups.
Best Regards,Donald McLaren, PhD
On Wed, Aug 3, 2016 at 8:35 AM, Marek Wypych <[log in to unmask]> wrote:
Dear SPMers,
We performed an experiment on readers and pre-readers with speech and print conditions. On second level analysis we computed null conjunction for the 2 contrast in each group (one sample t-test) which produced significant activation in readers while not in pre-readers. However, when the paper was sent to review one of the reviewers wanted us “to compute the conjunction of print and speech on the first level and put these contrasts into a 2-sample t-test on the second-level”.
However we came across some difficulties. If we save thresholded SPM map from each subject when performing null conjunction in SPM contrast manager and feed these maps to 2nd level the model cannot be estimated and there is an error saying “there are no inmask voxels” (even though we used very lenient p<0.05 uncorrected threshold for the 1st level conjunction). Is there a better way of doing multisubject conjunction?
Thank you in advance,
Marek
-- Marek Wypych PhD Laboratory of Brain Imaging (LOBI) Neurobiology Center Nencki Institute of Experimental Biology Pasteur 3, 02-093 Warsaw, Poland Tel.: +48 22 5892 550 http://lobi.nencki.gov.pl/
-- Marek Wypych PhD Laboratory of Brain Imaging (LOBI) Neurobiology Center Nencki Institute of Experimental Biology Pasteur 3, 02-093 Warsaw, Poland Tel.: +48 22 5892 550 http://lobi.nencki.gov.pl/