<div dir="ltr"><div><div><div>Dear Fieldtripers,<br><br>Following Maris et al. 2007 paper about "Nonparametric testing of coherence
difference" I was hesitating regarding the number of degrees of freedom to use in the Z-coherence transform when coherence is calculated over N trials by
averaging across M non-overlapping windows in each of them. Is the df equals to N or N*M then?<br><br>I am a little confused since in Fieldtrip, this is even 2*N that appears to be used: <a href="http://fieldtrip.googlecode.com/svn/trunk/statfun/ft_statfun_indepsamplesZcoh.m" target="_blank">http://fieldtrip.googlecode.com/svn/trunk/statfun/ft_statfun_indepsamplesZcoh.m</a><br><br></div>Thanks for your time.<br><br></div>Best,<br><br></div>Guillaume<br></div>