Dear CONN users, 

I'm having some difficulties trying to 'speed-up' post-hoc t-testing after doing my original interaction analyses. 

Some background information: 

My current setup aims to investigate whether active treatment vs. sham influences FC differently, and moreover, if there are differences in these FC changes between two groups (trait_low, trait_high) -> group x condition x time interaction. 
In a crossover design, all participants received as well an active treatment as a sham treatment.
 
Between-subject contrast for age(centered), trait_high, trait_low = [0 1 -1]
Between-conditions (condition x time interaction) contrast for Treatment_pre, Treatment_post, Treatment_post2, sham_pre, sham_post, sham_post2 = [1 -1 0 -1 1 0; 0 1 -1 0 -1 1]

When running this 2nd level ROI-to-ROI analysis in result explorer, some connections remained significant. 

If, according to the help forum, I want to make post-hoc testing somewhat more automatized instead of doing it visually and separately for each t-test, it was suggested to run the original analysis and 'export/save a mask' of the significant connections in the result explorer. Then you could perform a new analysis but this time leaving the default (eye(x)) contrasts in the between-subjects and between-conditions fields. After which some other steps should take place... 

However, I cannot find this 'export/save mask' option in the result explorer (only the connectome ring, stats, etc.). 

What I can do, is re-running the analysis with the [0 1 0; 0 0 1] between subject contrast, and the eye(6) between-conditions contrast, select a significant connection (i.e., seed and target, based on the original interaction analysis I did before) and then "plot effects" in a bar graph (containing 12 bars then) in the main GUI 2nd level results window.

However, for this bar graph, I want to see where my significant effects are (t test statistics, effect-sizes, p-values etc.) for all possible t-tests. 

Am I missing the 'export/save mask' option somewhere? (Conn 17.e)
Or is there another way to "speed-up" post-hoc t-testing? 

Any help is kindly appreciated! 

All the best, 
Josefien