Dear Helmut,
thank you for your response, you've been extremely helpful! You saved me a lot of trouble! Thank goodness I haven't still published any result of the analysis I'm doing!
So..
>> is SPM miscomputing degrees of freedom?
>Not really, see above. However, SPM relies on the dfs of the whole model, independent of the contrast. E.g. I have seen people running a two-sample t-test and reporting [1 0], [-1 0] as (de)activations for group 1. This way, you have too many df of course (those from group 2).
Well, then in a kind of sense it is miscomputing DOFs. I assumed (don't know why, it's just not mentioned everywhere) that when I performed second level analysis, where contrasts of the kind [1 0] might be of some interest, doing [1 0] produced valid maps! Also, since currently I'm running an ANOVA model where I also run post-hoc, I used to believe that performing subsequent tests from the same design Matrix was correct. Indeed it appears it totally isn't, since I'm (wildly) overestimating DOFs, therefore inflating t-values.
Therefore I should run a separate SPM analysis for each post-hoc. I'll keep this in mind.
To give an example,
>Well, in principle one should adjust the models accordingly then / set up separate models for those tests. The same holds for post-hoc tests, they should be based on the correct error terms / dfs, but I doubt there's an ultimate answer (at least this was my impression when looking at different textbooks focusing on statistics in general, lots of different suggestions).
I have a 4 level ANOVA. I make my F contrast: [1 -1 0 0;0 1 -1 0; 0 0 1 -1] and find some significant differences. Than I "assume" that the culprit is the first column and I want to run a second ANOVA without it. Now I was just defining a new contrast [0 1 -1 0; 0 0 1 -1]. Apparently doing this overestimates DOF including also those from the first column.
I'll just run a full separate analysis each time then!
Thank you a lot!
Luca
Rispetta l’ambiente: non stampare questa mail se non è necessario.
Respect the environment: if it's not necessary, don't print this mail.
|