 |
 |
Dear experts, I am analyzing data from a study with 2X2 design (2 groups [G1, G2], 2 conditions [MJ, NT]). The data were collected in a single run with conditions in a block design. My steps were as follows: 1. One first-level FEAT analysis per subject with 2 EV's (MJ, NT) and 4 contrasts each to model effects of condition (MJ, NT, MJ>NT, NT>MJ) 2. Group level analysis with 2 EV's (G1, G2) and 4 contrasts to model effects of group (G1, G2, G1>G2, G2>G1) I carried up all 4 lower level contrasts to the higher level analysis. I am most interested in the MJ>NT lower-level contrast, but I am confused with the results at the group level. The group level analysis shows 3 significant clusters for G1 and 0 significant clusters (no significant activation) for G2, but the G1>G2 higher-level contrast shows no significant activation. Is there an explanation for this, or are these results likely invalid for some reason? Any insight is appreciated, Laura