Abstract
Methods of correcting for continuity in tests of significance of the difference between correlated proportions are presented. These corrections should increase the range of usefulness of the formulas developed by McNemar (1).
Users
Please
log in to take part in the discussion (add own reviews or comments).