5.3. Exact Tests for Association
5.3.1. General Association
In some cases, there is not sufficient sample size for the chi-square statistics discussed earlier in this chapter to be valid (several mij ≤ 5). An alternative strategy for these situations is the Fisher exact test for s × r tables. This method follows the same principles as Fisher’s exact test for the 2 × 2 table, except that the probabilities that are summed are taken from the multivariate hypergeometric distribution. Mehta and Patel (1983) describe a network algorithm for obtaining exact p-values that works much faster and more efficiently than direct enumeration; Baglivo, Olivier, and Pagano (1988), Cox and Plackett (1980), and Pagano and Halvorsen (1981) have also done work in ...
Get Categorical Data Analysis Using The SAS® System, 2nd Edition now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
