Hello, 
I have multiple equivalent contrasts for the same studies. How can I take into account several contrasts for one study given the dependencies ? I read in the forum another topic about dealing with multiple contrasts but it was for images. 

How can I combine multiple contrasts for one study in the case of coordinates tables ?

Thanks !

Layla

Hello Layla,

The best approach to handle with multiple contrast for the same study, is to consider them as separate studies with specific and explanatory names. For each contrast, you should report the cordinates in the corresponding .txt file with the specific label.

Kind regards,

Lydia

Hello Lydia, 

Thank you for your answer. I see 3 possibilities in case of multiple contrasts: considering them as separate studies, combining them or choosing the most pertinent to answer the meta-analysis question.
I have included studies with multiple contrasts (evaluating the same mechanism) for the same population. I guess it imply intra-study dependency between the results obtained for each contrast. In this particular case, don't you think that considering each contrast as an independent study could represent an important bias in my CBMA (for instance, inflating the true effect)?
 
As far as I understand the previous responses of developers to similar questions, combining different contrasts from the same study as not yet been implemented in the new version of SDM. For now, I choose to include only the most pertinent contrast from each study displaying multiple contrast. What do you think of this approach?

Best,

Layla

Hello Layla,

I've just answered the other question you posted; explained better. I thought the subjects were not exactly the same so I posted a wrong answer. You are right that including all as different studies would biased the results in this case. The best approach then is to choose just one contrast, as you just said.

Best, 
Lydia

Hello Lydia,
Thank you for your answer ! I will definitely choose the most pertinent contrasts.

Best, 
Layla