I posted a question about calculating partial area under the curve on
the listserv a few days ago. I was pointed in several informative
directions but I think my question did not have enough information
concerning my problem. I am not calculating AUCs for indiviudals at
multiple time points. I have a large data file from which I have
generated individual ROC curves for various dependent measures
that are either individual dependent variables(such as cervical length)
or derived variables from various regression models. What I need to do
now is compare partial AUCs for these various measures
using only the portion of the ROC curve below a defined (1-Specificity)
or False Positive Rate value.
So I am only suing the ROCs generated in SPSS ant their corresponding
coordinates for Sensitivity and (1-Specificity) to calculate the partial
AUC and their std. errors in ordeer to make comparisons between the
various dependent measures. I don't know if I can do this with SPSS or
some other available software or equations. Any help would be appreciated.
--
MICHAEL KRUGER
Wayne State Univ. School of Medicine
Dept. of OB/GYN
"A TRUE PRINCE"
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
