Dear SPMers,
I'm struggling with a model estimation issue that seems to have come up multiple times in the archives, yet I am unable to find a documented solution or explanation. I am running a two-sample t-test in SPM8. When I estimate the model, it stops midway and I get the following MATLAB warning:
Initialising parameters : ...done
Plane 69/69 , block 1/1 : ...done
Warning: Please check your data: There are no significant voxels.
> In spm_spm at 815
In spm_run_fmri_est at 33
In matlabbatch/private/cfg_run_cm at 29
In cfg_util>local_runcj at 1477
In cfg_util at 803
In cfg_ui>MenuFileRun_Callback at 1218
In gui_mainfcn at 75
In cfg_ui at 53
Failed 'Model estimation'
Reference to non-existent field 'Vbeta'.
In file "/usr/local/pkg/spm8/config/spm_run_fmri_est.m" (v3691), function "spm_run_fmri_est" at line 100.
If I run the same model in SPM5, I get this message:
Error running job:
Error using ==> plot
Vectors must be the same lengths.
Interestingly, I have used these exact same contrast files in a 2x2 factorial with no problem at all. To troubleshoot, I have checked the level 2 t-test model to make sure all of the inputs and settings are correct. I have checked each subject's contrast images, and they all appear fine. I have also gone through and tried to run the model again subtracting out each subject one by one, and cant find any "bad" subjects that are responsible for the failure. I have also installed the latest updates to SPM8 in case it is a program bug. I would be tremendously appreciative if anyone can shed light on this issue. Many thanks in advance, Ben
|