Dear Mark,
I've just done a test analysis on my version of SPM:
Basic models
One Way ANOVA
Conditions [6]
- i then entered 11 images per condition ie. 11 subjects
Grand Mean Scaling [No]
threshold masking [none]
Implicit mask [yes]
Explicit mask [no]
Global [omit]
Non-sphericity corr [Yes]
errors are identical [No]
Unequal variances are between [conditions]
errors are independent [no]
ie. a replication of yours.
And it works !
Yours may not work because
a) you don't have the same version of SPM as I do.
you may have an earlier version. Are you asked different
questions about sphericity ? - indeed you may have given
inappropriate responses as to what the non-sphericity was
'over' - understandable, as these questions were somewhat obscure !
- so - upgrade by downloading the latest patches from
ftp://ftp.fil.ion.ucl.ac.uk/spm/spm2_updates/
b) don't know.
If you're really rushed to get something working for an HBM abstract
then try (a) or assume sphericity !
Best,
Will.
Mark Schram Christensen wrote:
> Dear Dr. Friston and Dr. Penny
> Today I posted a question to the SPM mailing list, and it seems from the
> answer that I received, that there is some kind of error in SPM2 when you
> are performing 2nd level ANOVA analysis using multiple groups (in this case
> 6) and non-sphericity correction.
>
> I am trying to run a 2nd level ANOVA of a number of F-contrasts using SPM2.
> From
> previous posts at the SPM-list, I can see that one has to use the
> beta-images
> from the 1st level analysis and then define the F-
> contrast after having performed the 2nd level analysis.
> I have 6 different (groups) effects from 1st level that I am interested in
> from 11 subjects,
> these are parametric modulations (linear and quadratic) from 3 different
> perceptual categories.
> I am doing non-sphericity correction over replications, as suggested
> previously
> (http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0312&L=spm&O=D&F=&S=&P=118
> 66)
> However, when I run the estimation-process it fails. The hyperparameter
> estimation works fine, but when the parameter estimation starts an error
> occurs (see bottom).
>
> I have tried to run the same analysis using only 3 groups (the linear
> parametric modulation), and that works fine, but I would like to study both
> the linear and quadratic effects, e.g. an F-contrast like:
>
> 1 0 1 0 1 0
> 0 1 0 1 0 1
>
> I have tried to run the analysis on both a Window XP Pro system and a Linux
> system, and it fails in both cases.
>
> Today, at the SPM-list, a reply was posted by S. F. W. Neggers, who one year
> ago encountered the same problem, i.e. for some combinations of number of
> groups and replications it failed. He had come to the same conclusion that
> it seemed to be a bug in SPM, and he has stopped using 2nd level ANOVAs for
> F-contrasts.
>
> If the solution is trivial it would come in very handy, if it was solved
> before the HBM abstract deadline. Or if any other solutions to come about
> the problem exist, I would appreciate to hear about it.
>
> Yours sincerely,
>
> Mark S. Christensen
>
> _________________________________________
> Mark Schram Christensen
> M.Sc., Ph.D. student
> Danish Research Centre for Magnetic Resonance
> Copenhagen University Hospital Hvidovre
> Kettegård Allé 30
> DK-2650 Hvidovre
> &
> Institute of Physical Exercise and Sport Science
> Copenhagen University
> Denmark
> Phone: +45 3632 3323/+45 2695 6275
> Fax: +45 3647 0302
> E-mail: [log in to unmask]
> Web: www.drcmr.dk
> www.ifi.ku.dk
>
>
> SPM2: spm_spm (v2.66) 12:02:01 - 07/01/2005
> ========================================================================
> Initialising parameters : ...computing???
> Index exceeds matrix dimensions.
>
> Error in ==> C:\MATLAB6p5\toolbox\matlab\sparfun\spdiags.m
> On line 102 ==> a = [a; i i+d(k) B(i+(m>=n)*d(k),k)];
>
> Error in ==> C:\MATLAB6p5\spm2\spm_spm.m
> On line 419 ==> s = spdiags(1./sqrt(diag
> (s)),0,nScan,nScan);
>
> Error in ==> C:\MATLAB6p5\spm2\spm_spm.m
> On line 826 ==> SPM = spm_spm(SPM);
>
> ??? Error while evaluating uicontrol Callback.
>
>
>
--
William D. Penny
Wellcome Department of Imaging Neuroscience
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7475
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
|