Calculating sensitivity, specificity, etc.of a diagnostic test

[George76]

I would like to calculate the specifity, sensitivity, positive and negative predictive value etc. for a diagnostic test, possibly with a method that uses p-values (sorry, i'm a newbie). I have coded my variable with 0/1 (o= disease not detected; 1= disease detected). The underlying disease is also coded with 0/1, of course. Is it possible to analyze whether disease detection changes in different subgroups of people (i.e. age > 70 vs < 70, male vs. females ecc.)? How?
Thank you very much in advance!!!

 

[TheEcologist]

Is it possible to analyze whether disease detection changes in different subgroups of people (i.e. age > 70 vs < 70, male vs. females ecc.)? How?
Thank you very much in advance!!!

I advise you to start reading up on logistic regression, that being the preferred tool for your problem.

Here's the perks in a nutshell:

http://statconsultant.********.com/2007/09/logistic-regression-in-nutshell.html

change the asterisks in the link to bl-og-sp-ot (and remove the -). The forums auto censure wont let me post that word.