MATH 203 Lecture Notes - Lecture 24: Contingency Table, Categorical Variable

• School: McGill University
• Department: Mathematics & Statistics (Sci)
• Course: MATH 203
• Professor: Michael Wallace

Document Summary

Categorical data experimental units are allocated in a number of categories: categories can have 1<= factors, e. g. ) gender/smoking/age groups. Even numerical data may be treated categorically. Data counts of experimental units in each category: n e. u"s, k categories, probabiliies of the k outcomes: p1, , pk. P1+ +pk = 1: e. u. "s are independent, the counts in the k categories are n1, ,nk. Mulinomial experiment the one with several types of counts: categories can arise as combinaions of factor levels. One-way (categories of a single factor) H0: p1 = p2 = pk = 1/k. Tests about the factors a and b. Are a and b dependent: i. e. does a classiicaion inluence b, be careful about using the words as. Inluence never make a causal statement. H0: p1 = p1(0), pk = pk(0) H0: p1 = p2 = p3 = 1/3. H0: p1 = , p2=p3 = .