Dear reader,
I have a question about how to facilitate the interpretation of a three-way interaction which I found with the following design:
On the subject level, there is one within subjects factor with three levels (negative, neutral, positive) for which 3 EV's are made. These EV's are modulated (weigthed) by the participants responses (coined applicability) rated on a scale of 1 to 4 contrast coded to -3, -1, 1, 3. On the group level one centered EV (beside the constant) is used (coined self-esteem). Higher level contrasts are set up to look at the effect of self-esteem on activation during the these three conditions modulated by applicability.
There is a three-way interaction between negative condition, applicability and self-esteem in the positive direction. Indicating that as self-esteem increases the activation in this area is more strongly related to applicability in the negative condition. I would like to visualise this relation and would therefore need to extract the relation between activation and applicability for the negative trials. However, I have not been able to find out how to this using e.g. featquery. I could use the time series from featquery, however, this returns the 'activation' per volume in this region where I would only need the activation during negative trials (which consists of multiple volumes) and the corresponding modulation.
Would you have any suggestions on how to obtain this?
Best, Charlotte
|