Hi,
In addition to determining significance values between netmats in different conditions (i.e. N subjects in condition A vs. a different N subjects in condition B), I would like to determine significance for condition A and condition B on their own. I realize that when nets_groupmean is used to calculate a Znet, it does a (parametric) one group t-test before conversion to z-statistics, and should in theory tell me about significance of edges, however it doesn't really provide the kind of output I was looking for; specifically what nets_glm, nets_edgepics and nets_boxplots provide.
Is there then value in doing a 1-sample t-test using nets_glm? If so, it clearly makes non-parametric assumptions, whereas nets_groupmean makes parametric assumptions... I was hoping you could help clarify which I should use when trying to find significance for single conditions on their own, and why there might or might not be an advantage to using non-parametric assumptions with nets_glm vs. making parametric assumptions with nets_groupmean.
Thanks very much for your help!
|