I have used Robert Centors ROC analyzer for
calculating the non-parametric ROC area of even binary
diagnostic values.
The ROC area is useful when comparing the
discriminating power of diagnostic variables
independent of the incidence of the disease, even for
binary variables.
I think this reference can be of interest:
The Area under an ROC Curve with Limited Information
Wilbert B. van den Hout
Roland Andersson
Ronan M Conroy wrote: on 19/12/2003 08:32,
fmedeiros at fmedeiros@u... wrote: Is there a minimum
number of categories to adequately fit a parametric
ROC curve for ordinal data? For example, can we
adequately fit a model when we have only 3 categories
(for instance, abnormal, borderline and normal)
You could try -nproc- by Philip Price and Fred Wolfe.
But with so few categories, you would be better off
looking at the performance of individual cutoff points
on the scale in terms of sensitivity, specificity,
positive and negative predictive values. The ROC area
is a very broad brush, and does not assess the
suitability of a test for specific purposes.
H�strusk och gr� moln - k�p en resa till solen p� Yahoo! Resor p� adressen http://se.docs.yahoo.com/travel/index.html
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
