help
86 Subscribers 
help > 2-group connectivity changes
Showing 1-3 of 3 posts
Oct 21, 2014 05:10 AM | Ravnoor Gill - University of Western Ontario
2-group connectivity changes
Dear Andrew,
I am trying to compare 2 groups for changes in connectivity. When I run the analysis with either contrast, [1,-1] or [-1,1], I get exactly the same set of networks with the same number of nodes and edges with the contrast using F-test with NBS. Same goes for FDR, however, the number of nodes and edges are significantly reduced.
I am using threshold 6 while having tried lower values. The connectivity matrix values are Fischer-Z transformed.
Any suggestions as to interpret/resolve such an outcome would be greatly appreciated.
Thank you!
Sincerely,
Ravnoor
Design Matrix:
I am trying to compare 2 groups for changes in connectivity. When I run the analysis with either contrast, [1,-1] or [-1,1], I get exactly the same set of networks with the same number of nodes and edges with the contrast using F-test with NBS. Same goes for FDR, however, the number of nodes and edges are significantly reduced.
I am using threshold 6 while having tried lower values. The connectivity matrix values are Fischer-Z transformed.
Any suggestions as to interpret/resolve such an outcome would be greatly appreciated.
Thank you!
Sincerely,
Ravnoor
Design Matrix:
0 1
0 1
0 1
0 1
0 1
1 0
1 0
0 1
0 1
1 0
1 0
1 0
0 1
1 0
0 1
0 1
0 1
0 1
1 0
1 0
0 1
0 1
1 0
1 0
1 0
0 1
1 0
Oct 21, 2014 07:10 AM | Andrew Zalesky
RE: 2-group connectivity changes
Hi Gill, it does not really make sense to use an F-test with a
contrast of [-1 1].
I suggest using a t-test with a contrast of [-1, 1] and/or [1, -1]. It is impossible for these two contrasts to give the exact same result, unless the statistic threshold is zero. Use a statistic threshold of 2 or 3 perhaps.
Andrew
Originally posted by Ravnoor Gill:
I suggest using a t-test with a contrast of [-1, 1] and/or [1, -1]. It is impossible for these two contrasts to give the exact same result, unless the statistic threshold is zero. Use a statistic threshold of 2 or 3 perhaps.
Andrew
Originally posted by Ravnoor Gill:
Dear Andrew,
I am trying to compare 2 groups for changes in connectivity. When I run the analysis with either contrast, [1,-1] or [-1,1], I get exactly the same set of networks with the same number of nodes and edges with the contrast using F-test with NBS. Same goes for FDR, however, the number of nodes and edges are significantly reduced.
I am using threshold 6 while having tried lower values. The connectivity matrix values are Fischer-Z transformed.
Any suggestions as to interpret/resolve such an outcome would be greatly appreciated.
Thank you!
Sincerely,
Ravnoor
Design Matrix:
I am trying to compare 2 groups for changes in connectivity. When I run the analysis with either contrast, [1,-1] or [-1,1], I get exactly the same set of networks with the same number of nodes and edges with the contrast using F-test with NBS. Same goes for FDR, however, the number of nodes and edges are significantly reduced.
I am using threshold 6 while having tried lower values. The connectivity matrix values are Fischer-Z transformed.
Any suggestions as to interpret/resolve such an outcome would be greatly appreciated.
Thank you!
Sincerely,
Ravnoor
Design Matrix:
0 1
0 1
0 1
0 1
0 1
1 0
1 0
0 1
0 1
1 0
1 0
1 0
0 1
1 0
0 1
0 1
0 1
0 1
1 0
1 0
0 1
0 1
1 0
1 0
1 0
0 1
1 0
Oct 24, 2014 02:10 PM | Ravnoor Gill - University of Western Ontario
RE: 2-group connectivity changes
Make sense, and works like a charm.
Thanks, Andrew!
Thanks, Andrew!