Displaying 1 result from an estimated 1 matches for "gsens".
Did you mean:
sens
1998 Feb 27
1
R-beta: is there a way to get rid of loop?
...dwards's G_I
(Likelihood, p. 194).
I label the cells in the table as follows
stim response
"y" "n" total
--------------------------------
y hit miss nsignal
--------------------------------
n false correct nnoise
alarm rejection
---------------------------------
Gsens<-function(nhit,nmiss,nfa,ncr)
{
# Edwards's G_I, measure of interaction, is measure of sensitivity
(xlogx(nhit)+xlogx(nmiss)+xlogx(nfa)+xlogx(ncr))-(xlogx(nhit+nmiss)+xlogx(nfa+ncr))-(xlogx(nhit+nfa)+xlogx(nmiss+ncr))+xlogx(nhit+nmiss+nfa+ncr)
}
BTW I can't figure out how to break the l...