[Rd] mcnemar.test helpfile (PR#9220)

candrews at buffalo.edu candrews at buffalo.edu
Tue Sep 12 16:17:33 CEST 2006

Full_Name: Chris Andrews
Version: 2.3.1
OS: Windows
Submission from: (NULL) (


ends with an example:

     ## Agresti (1990), p. 350.
     ## Presidential Approval Ratings.
     ##  Approval of the President's performance in office in two surveys,
     ##  one month apart, for a random sample of 1600 voting-age Americans.
     Performance <-
     matrix(c(794, 86, 150, 570),
            nr = 2,
            dimnames = list("1st Survey" = c("Approve", "Disapprove"),
                            "2nd Survey" = c("Approve", "Disapprove")))
     ## => very strong association between the two successive ratings

The example runs fine.  The final statement ("very strong association between
the two successive ratings") is true.  But they are unrelated.  McNemar's test
is a test of marginal homogeneity.  The conclusion should be that opinion has
changed.  The marginal distributions are different.  The approval rating has
dropped from 59% to 55% (p=4e-5).

More information about the R-devel mailing list