open-discussion
open-discussion > RE: Need help
Sep 3, 2011 11:09 AM | Ged Ridgway
RE: Need help
Dear Ramesh,
1. I'm afraid your results are not significant following correction for either FWE or FDR, so you shouldn't over-interpret the uncorrected p<0.001. See e.g.
http://www.ajnr.org/cgi/content/full/31/... and ref.20 there
2. & 3. A possible explanation is that the results are in MNI space, not Talairach, see e.g.
http://imaging.mrc-cbu.cam.ac.uk/imaging...
http://brainmap.org/icbm2tal/index.html
The other factor is that you have probably not used a very strict mask, as your search volume is about 1.8 litres, which seems a bit large. If you are using absolute or relative threshold masking, try increasing it, or try making an average of all the images (the smoothed normalised ones that you are doing the stats on) and binarising it with a threshold around 0.2, both using ImCalc. A stricter mask will make your FWE p-values more significant, though I don't expect it will change them enough to get close to 0.05.
I will email you PDFs for the Barnes and Henley articles separately.
Best,
Ged
1. I'm afraid your results are not significant following correction for either FWE or FDR, so you shouldn't over-interpret the uncorrected p<0.001. See e.g.
http://www.ajnr.org/cgi/content/full/31/... and ref.20 there
2. & 3. A possible explanation is that the results are in MNI space, not Talairach, see e.g.
http://imaging.mrc-cbu.cam.ac.uk/imaging...
http://brainmap.org/icbm2tal/index.html
The other factor is that you have probably not used a very strict mask, as your search volume is about 1.8 litres, which seems a bit large. If you are using absolute or relative threshold masking, try increasing it, or try making an average of all the images (the smoothed normalised ones that you are doing the stats on) and binarising it with a threshold around 0.2, both using ImCalc. A stricter mask will make your FWE p-values more significant, though I don't expect it will change them enough to get close to 0.05.
I will email you PDFs for the Barnes and Henley articles separately.
Best,
Ged
Threaded View
| Title | Author | Date |
|---|---|---|
| Ramesh Babu MG | Aug 31, 2011 | |
| Mark Jason | Nov 11, 2011 | |
| Ramesh Babu MG | Nov 15, 2011 | |
| Ged Ridgway | Sep 1, 2011 | |
| Ramesh Babu MG | Sep 3, 2011 | |
| Ged Ridgway | Sep 3, 2011 | |
| Ramesh Babu MG | Sep 4, 2011 | |
| Ged Ridgway | Sep 5, 2011 | |
| Ramesh Babu MG | Sep 5, 2011 | |
| Ramesh Babu MG | Sep 1, 2011 | |
| Cartik Sharma | Aug 31, 2011 | |
| Ramesh Babu MG | Sep 1, 2011 | |
| Cartik Sharma | Sep 1, 2011 | |
| Ramesh Babu MG | Aug 31, 2011 | |