Dear Adrian,
Your t-contrasts are correct and test for an effect of each of your 4 conditions compared to the mean of the remaining conditions. Your F-contrast tests for 'any difference' between conditions of both signs (+ or -). Multiplying your t-contrasts by -1 simply reverses the sign and you would test for a positive effect of the mean of 3 conditions compared to the mean of 1 condition.
Martin
--
Martin Dietz
[log in to unmask]
On Nov 25, 2010, at 9:51 AM, Adrian Imfeld wrote:
> Dear SPM experts,
>
> I'd like to do a 2nd level one-way ANOVA with a within subject factor, as introduced with the SPM8 updates 4010. I have 11 subjects and a 4 experimental conditions. I can set up the model just fine but I'm not sure if I'm doing the contrasts right:
>
> I noted that for this type of model, SPM accepts only contrasts which sum up to zero. To compare each of the conditions against the other ones I use 4 separate t-contrasts:
> 1: 0 0 0 0 0 0 0 0 0 0 0 3 -1 -1 -1
> 2: 0 0 0 0 0 0 0 0 0 0 0 -1 3 -1 -1
> 3: 0 0 0 0 0 0 0 0 0 0 0 -1 -1 3 -1
> 4: 0 0 0 0 0 0 0 0 0 0 0 -1 -1 -1 3
>
> For the main effect over all 4 conditions, I specify an F-contrast containing the above 4 contrast rows.
>
> So here are my questions:
> 1. Are the above contrasts correct.
> 2. Why do the contrasts have to sum up to 0? I thought the F-contrast should be [zeros(4, 11) eye(4)], but SPM says "invalid contrast".
> 3. Can I simply multiply my contrasts by -1 to test for "negative" effects?
>
> Thanks for your help, Adrian
|