Sample size for the kappa-statistic measure of interrater agreement (command-line version) ------------------------------------------------------------------- See help for @sskdlg@ for an in-depth description. ^sskapp , p1() p2() diff() kapp() [ n() level()]^
^p1()^ is a proportion (prevalence) between 0 and 1 ^p2()^ is a proportion (prevalence) between 0 and 1 ^diff()^ is the width of the confidence interval ^kapp()^ is the kappa for which you want the sample size
Options ------------ ^n()^ will give you the difference you can find at a given level
Returns ------------ ^r(nfixd)^ gives you the sample size for a fixed d, p1,p2 and kappa. ^r(dfixn)^ gives you the difference for a fixed n, p1,p2 and kappa.
Author ------ @sskdlg@ was written by Michael E. Reichenheim, IMS/UERJ, Brazil, michael@@ims.uerj.br The command-line version was modified from @sskdlg@ by Jan Brogger, University of Bergen, Norway, jan.brogger@med.uib.no
Also see --------
STB: STB-58: sxd3 On-line: help for @kap@, @kappa@