Categorical Data Analysis Using The SAS® System, 2nd Edition

4.2. Sets of 2 × r Tables

Consider the data from Koch and Edwards (1988) displayed in Table 4.1. The information comes from a randomized, double-blind clinical trial investigating a new treatment for rheumatoid arthritis. Investigators compared the new treatment with a placebo; the response measured was whether there was no, some, or marked improvement in the symptoms of rheumatoid arthritis.

Table 4.1. Rheumatoid Arthritis Data

		Improvement
Gender	Treatment	None	Some	Marked	Total
Female	Test Drug	6	5	16	27
Female	Placebo	19	7	6	32
Total		25	12	22	59
Male	Test Drug	7	2	5	14
Male	Placebo	10	0	1	11
Total		17	2	6	25

These data comprise a set of two 2 × 3 tables. There is interest in the association between treatment and degree of improvement, adjusting for gender effects. Degree of ...