open-discussion > interpretation of second-level analysis results
Showing 1-1 of 1 posts
Display:
Results per page:
Jul 9, 2019  11:07 AM | Renxi Li
interpretation of second-level analysis results
Dear author:

I am a researcher on the connectivity of Parkinson's disease patients. In our study, we recruited PD (Parkinson disease patients), controls as our groups. Further, we subdivide PD into two subgroups - AR (akinetic rigid group) and TR (tremor-dominate group). We are now in the 2nd-level analysis. 

Here are my question:

1. We have several covariates, such as age, gender for PD and controls. When comparing the PD subgroups (e.g. AR vs TR), do we need to make covariates for the subgroups (e.g. specify AR as 1, all others as 0), or we can use the covariates we defined for PD and controls?

2. For the between-subjects contrast, what does average [0.5 0.5] mean? What does the effect size of its results mean? Can we add covariates in this contrast?

3. For the between-subjects contrast, when doing F test, it there a way to do post-hoc testing?


Best,
Renxi