Dear Min,

In SPM8 FDR topological inference was introduced, and voxel-wise FDR inference hidden.  Topological inference means inference on peaks and clusters; voxel-wise inference is based on every individual voxel in the image (instead of spatial features of the image).  Thus "Topological FDR" means inference on clusters based on cluster size (or local peaks based on peak height), controlling the fraction of false positive clusters among all clusters (or false positive peaks among all peaks) on average, over many experiments.

While topological FDR results may be easier to interpret, in my experience it is is generally not very sensitivity and yields similar results to FWE-corrected inferences.  This would explain your reduction in sensitivity moving from SPM2 to SPM8.

If you would like to use voxel-wise FDR in SPM8, edit spm_defaults, changing "topoFDR" line to read
defaults.stats.topoFDR      = 0;
(quit and re-start SPM to take effect).

For more on this see references below.

-Tom

Chumbley, J., Worsley, K., Flandin, G., & Friston, K. (2010). Topological FDR for neuroimaging. NeuroImage, 49(4), 3057-64. doi: 10.1016/j.neuroimage.2009.10.090.

Chumbley, J. R., & Friston, K. J. (2009). False discovery rate revisited: FDR and topological inference using Gaussian random fields. Neuroimage, 44(1), 62–70. doi: 10.1016/j.neuroimage.2008.05.021.


On Tue, May 10, 2011 at 11:27 AM, Min <[log in to unmask]> wrote:
Dear SPM's users,

I ran spm2 and spm8, respectively, for the same set of fMRI data.  However, though I followed the same steps for the two versions of spm, there are considerable differences in the results.  With spm8, no suprathreshold clusters were found with fdr(0.05) corrected, while with spm 2, more than 8 clusters survive the threshold (T values of height threshold are nearly the same for the two analysis).

I am unsure how to explain such a big difference between the results from two versions of spm.

I will be grateful for any help!

Sincerely,
Min



--
____________________________________________
Thomas Nichols, PhD
Principal Research Fellow, Head of Neuroimaging Statistics
Department of Statistics & Warwick Manufacturing Group
University of Warwick
Coventry  CV4 7AL
United Kingdom

Email: [log in to unmask]
Phone, Stats: +44 24761 51086, WMG: +44 24761 50752
Fax:  +44 24 7652 4532