help
help > RE: t-test and p-value in graph theory
Jan 27, 2017 01:01 AM | Alfonso Nieto-Castanon - Boston University
RE: t-test and p-value in graph theory
Hi,
Assuming that you are performing a two-sample t-test (e.g. you selected two second-level covariates identifying two different subject groups and entered a between-subjects contrast [1 -1]) then the top line in the results table of the graph theory window shows you whether the selected measure (e.g. global efficiency) at the network-level (e.g. network global efficiency, defined as the average global efficiency across all nodes of the network) is different between the groups (this is a two-sample t-test of the network global efficiency measures across subjects). A not-significant value there (e.g. p>.05) means that network global efficiency does not differ between the groups. That does not preclude global efficiency at individual nodes from perhaps being different between the two groups. You can identify potential nodes with different global efficiency between the two groups by looking in that same table whether any individual node/ROI result shows p-FDR<.05 (or by setting the corresponding statistical threshold in the gui, and seeing whether any ROI survives that correction)
Hope this helps
Alfonso
Originally posted by snowa wo:
Assuming that you are performing a two-sample t-test (e.g. you selected two second-level covariates identifying two different subject groups and entered a between-subjects contrast [1 -1]) then the top line in the results table of the graph theory window shows you whether the selected measure (e.g. global efficiency) at the network-level (e.g. network global efficiency, defined as the average global efficiency across all nodes of the network) is different between the groups (this is a two-sample t-test of the network global efficiency measures across subjects). A not-significant value there (e.g. p>.05) means that network global efficiency does not differ between the groups. That does not preclude global efficiency at individual nodes from perhaps being different between the two groups. You can identify potential nodes with different global efficiency between the two groups by looking in that same table whether any individual node/ROI result shows p-FDR<.05 (or by setting the corresponding statistical threshold in the gui, and seeing whether any ROI survives that correction)
Hope this helps
Alfonso
Originally posted by snowa wo:
After enter a [1
-1] between subjects contrast in graph theroy,
p value at the top of the table is mean what?
Results of t-test between groups or some rois of the network?
If i got no-significant p-value(>0.05) at this place(uppermost of table),
can I judge the values of the roi below them to be meaningful?
It's really hard to understand.
so please answer me by all means.
best regards.
p value at the top of the table is mean what?
Results of t-test between groups or some rois of the network?
If i got no-significant p-value(>0.05) at this place(uppermost of table),
can I judge the values of the roi below them to be meaningful?
It's really hard to understand.
so please answer me by all means.
best regards.
Threaded View
| Title | Author | Date |
|---|---|---|
| snowa wo | Jan 25, 2017 | |
| Alfonso Nieto-Castanon | Jan 27, 2017 | |
| snowa wo | Jan 28, 2017 | |
| Pravesh Parekh | Jan 31, 2017 | |