> vote=matrix(c(794,86,150,570),nr=2,dimnames=list("1st Survey"=c("Approve","Disapprove"),"2nd Survey"=c("Approve","Disapprove"))) > vote 2nd Survey 1st Survey Approve Disapprove Approve 794 150 Disapprove 86 570 > ##Set correct=F to apply McNemar without continuity correction > mcnemar.test(vote,correct=F) McNemar's Chi-squared test data: vote McNemar's chi-squared = 17.3559, df = 1, p-value = 3.099e-05 > ##Set correct=t to apply McNemar with continuity correction > ##Note: we don't really need the correction here since the sample size is large > mcnemar.test(vote,correct=T) McNemar's Chi-squared test with continuity correction data: vote McNemar's chi-squared = 16.8178, df = 1, p-value = 4.115e-05 > > ### Simple Kappa Coefficient > > ### Using the original formula to calculate the Simple Kappa Coefficient > prop=vote/sum(vote) > Po=sum(diag(prop)) > Pe=rowSums(prop)[1]*colSums(prop)[1]+rowSums(prop)[2]*colSums(prop)[2] > kappa=(Po-Pe)/(1-Pe) > kappa Approve 0.6995927 > > ### Using a function Kappa() in package vcd. > ### Please first load packages VR, colorspace and grid and finally load vcd. > #install.packages("vcd") > library(vcd) > kappa=Kappa(vote) > CI_kappa=cbind(0.69959267-qnorm(0.975)*0.01805518,0.69959267+qnorm(0.975)*0.01805518) > CI_kappa [,1] [,2] [1,] 0.6642052 0.7349802 > ### or use confint() function > confint(kappa) Kappa lwr upr Unweighted 0.6642052 0.7349802 Weighted 0.6329452 0.7662402 > > ###agreement plot > ####observed and expected diagonal elements are represented by superposed black and white rectangles, respectively. > agreementplot(vote) > > > ####Consider cross-sectional design > vote=matrix(c(944,880,656,720),nr=2,dimnames=list(c("1st Survey","2nd Survey"),c("Approve","Disapprove"))) > vote Approve Disapprove 1st Survey 944 656 2nd Survey 880 720 > ##The ususal chi-sq. test > chisq.test(vote, correct=F) Pearson's Chi-squared test data: vote X-squared = 5.2224, df = 1, p-value = 0.0223 > ##Set correct=F to apply McNemar without continuity correction > mcnemar.test(vote,correct=F) McNemar's Chi-squared test data: vote McNemar's chi-squared = 32.6667, df = 1, p-value = 1.094e-08 > ##Set correct=t to apply McNemar with continuity correction > ##Note: we don't really need the correction here since the sample size is large > mcnemar.test(vote,correct=T) McNemar's Chi-squared test with continuity correction data: vote McNemar's chi-squared = 32.3757, df = 1, p-value = 1.271e-08