Dear Snpm experts,

I want to correlate 17 PET scans with corresponding behavioral score using the non parametric approach proposed in Snpm96, but I encounter some trouble with variance smoothing. 

I use snpm plug-in "SingleSub: Single Covariate of interest". After scans selected and covariate entered, I have the following messages (with my answers):

======================================================================
Size of exchangability block
======================================================================

======================================================================
	3.556874e+14 Perms. Use approx. test?
======================================================================

[y]es / [n]o ? y


======================================================================
	# perms. to use? (Max 3.556874e+14)
======================================================================

# perms. to use? (Max 3.556874e+14) : 4000

======================================================================
	FWHM(mm) for Variance smooth
======================================================================

FWHM(mm) for Variance smooth (Default: 0 ) : 10 10 10
WARNING: Working volumetrically because of smoothing in z (10),
but more than 17 scans analyzed.
May run outof memory.

======================================================================
	Collect Supra-Threshold stats?
======================================================================

[y]es / [n]o  (Default: no )? y


======================================================================
	Select global normalisation
======================================================================
	 1 : No Global Normalisation  
	 2 : Proportional scaling     
	 3 : AnCova                   

Menu choice (1-3) ? 2


======================================================================
	Gray matter threshold ?
======================================================================

Gray matter threshold ? (Default: 0.8 ) : 

======================================================================
	Value for grand mean ?
======================================================================

Value for grand mean ? (Default: 50 ) : 



And the file SnPM_cfg.mat is then created. The problem came in the second step. When I try to load this file, I get an error message :

??? Error using ==> spm_snpm
Cannot z-smooth variance in non-volumetric mode

Thus in the first step I was allowed to smooth variance but only under Working volumetrically condition (cfr the warning in the setup stage), and now it seems that the program don't recognize that these computations must be done in volumetric mode. 

I will be very grateful to everyone who may help me to understand this and how I can manage it. 

Best regards to all
    ////        ,,,,       ////
--[[. .]]------[* *]------[@ @]--------------
    [~]	    [~]	   [#]

PEIGNEUX Philippe, Lic. Psych., Chercheur

E-mail: [log in to unmask]

Cyclotron Research Centre - Neurology Unit
Liege University
Batiment B30 
Allee du 6 Aout, 8
B-4000 Liege
BELGIUM	
phone: +32-4-3662316
fax:   +32-4-3662946

Neuropsychology Department
Liege University
Batiment B33 						
Boulevard du Rectorat, 3				      
B-4000 Liege
BELGIUM						       
phone: +32-4-3662394     				   			   	    
fax:   +32-4-3662808				       

---------------------------------------------


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%