i would check out randy mcintosh's pls software
http://www.rotman-baycrest.on.ca/index.php?section=84
hth, ajs
On 4/16/08, Will Penny <[log in to unmask]> wrote:
> Martin Havlíček wrote:
> > Dear sir,
> >
> > I want to ask you if you can answer me some questions about SEM
> > modeling. I have read your article "Modelling Functional Integration: A
> > Comparison
> > of Structural Equation and Dynamic Causal Models" and there are few
> > things, which are not clear to me, because I am not so experienced in
> > statistic.
> >
> > I want to implement stacked model to compare two conditions of attention
> > and non-attention and want to also consider reciprocal path connection
> > (in same way as you have it in your article). But don't understand how
> > exactly is meant the comparison between null and alternative model. It
> > seems me that this comparison is already in equation:
> > F = log(|implicit|) + Tr(observed*inv(implicit)) - log(|observed|) - N.
> > Does it mean that I should just perform SEM modelling with this equation
> > for data partition that belongs to condition of attention and partition
> > that belongs to non-attention (i.e. do it twice) and then compare them
> > by chi-square difference? If I do it in this way, then I have no idea
> > how to satisfy the conditions that in null model path coefficients
> > should be equal between conditions (attention and non-attention) and in
> > alternative model can vary.
> >
>
> Yes, its as you describe.
>
> You have one data set D that is partitioned into two parts; attention D1
> and no attention D2.
>
> For the alternative model a connection of interest is allowed to vary
> between D1 and D2 and for the null model it is not.
>
> In the statistical comparison the data is then viewed as one data set D
> and the above equation used.
>
> The satisfaction of the constraints is something that your SEM software
> should handle for you during its optimization. You just tell it what the
> constraints are.
>
> SPM does not provide SEM modelling so I can't help you with any software
> tips here other than to say try the student edition of LISREL (but I've
> never used it !).
>
> Best wishes,
>
> Will.
>
> > I will be very grateful for any advice.
> >
> > Martin Havlicek.
> >
> > Btw. really appreciate your work in SPM team.
> >
> >
> >
> >
> >
> >
> >
> >
>
> --
> William D. Penny
> Wellcome Trust Centre for Neuroimaging
> 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/
>
--
Alexander J. Shackman
Laboratory for Affective Neuroscience
Waisman Laboratory for Brain Imaging & Behavior
University of Wisconsin-Madison
1202 West Johnson Street
Madison, Wisconsin 53706
Telephone: +1 (608) 358-5025
FAX: +1 (608) 265-2875
EMAIL: [log in to unmask]
http://psyphz.psych.wisc.edu/~shackman
Calendar {still under construction}:
http://www.google.com/calendar/embed?src=ajshackman%40gmail.com
|