I am trying to understand the stats that are incorporated within Higher-Level FEAT and would be grateful for any input - sorry for the basic question.

When I do single group averages of two Groups I can see that group has activation in more areas than Group 1. https://www.dropbox.com/sh/p73n1d0iwkv715u/AABbV6f8LM3-fdtHb-zYFqdIa?dl=0

However, now when I do higher level FEAT and select the option in the model setup for comparing two groups (unpaired), there is no significant difference between the groups (When both Group1>Group 2 and Group2>Group1 are tested). 

Can you please help me understand this result? Since there is activation in the right S1 in Group 2 but no activation at all in Group 1, how can there be no significant difference when stats are run? 

Thank you very very much

Sheena