Hello FSL List Serv,
I am curious about whether I am methodologically on track with my group level analysis of a Stroop Study.
Background on Stroop/Study:
The basic concept is outlined here: http://en.wikipedia.org/wiki/Stroop_effect. We are concerned with finding a stroop effect and conflict adaptation effect. Briefly, a Stroop effect is found by measuring the difference between congruent trials (when both of the stimuli are similar in nature) and incongruent trials (when the stimuli are of opposite nature). Conflict adaptation occurs when an incongruent trial is followed by another incongruent trial. The subjects respond quicker in this case than when an incongruent trial is preceded by a congruent trial.
Thus, we have two types of trials for the stroop effect: Congruent and Incongruent
And four types of trials for the conflict adaptation effect: Inc-Inc, Con-Inc, Con-Con, Inc-Con
Question:
The study has two scans with two conditions (one scan in each condition). Within each scan there are two separate runs of a Stroop study. We are concerned with finding a difference between conditions in stroop effect and conflict adaptation.
Proposed plan steps:
1) Do First Level analysis
2) Do a fixed effects for each subject, each condition for the two runs. To find the average of the two runs within each scan.
3) Do a fixed effects for each subject comparing the two conditions. We would use the average generated for each condition from step 2.
4) Do a mixed effects for a group level average of the comparison.
So that is my proposed design. I was wondering if anyone could offer advice as to whether this was the proper way to conduct it and if any improvements could be made.
Thank You,
Michael Cordell
|