 |
 |
Dr. John, Thank you for your input. Currently I am using 8 control and 8 KO mice. But people have reported the significance even with less than 8 animals. Sorry for my misunderstanding.
Thanks
Ned
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of John Ashburner
Sent: Tuesday, June 23, 2015 12:26 PM
To: [log in to unmask]
Subject: Re: [SPM] VBM results
I see a t test as a test of whether you have enough subjects to show a difference ("because the probability of an effect being zero is vanishingly small, given enough scans or subjects one can always demonstrate a significant effect at every voxel" KJF). With only 16 mice, you won't have a particularly powerful VBM analysis.
Best regards,
-John
On 23 June 2015 at 16:17, Ned <[log in to unmask]> wrote:
Hello SPM VBM experts
I tried to compare the differences between two mouse groups ( control vs knockout), I was able to do registration, normalization and segmentation nicely. I did 0.5 mm isotropic smoothing Then I did 2 sample t test on grey matter with p=0.01 with 200 clusters, I got the above results. I am wondering that why I am getting few clusters with significant voxels? Does this look normal? Please share your suggestions.
Thanks
Ned