STATS 2B03 Lecture Notes - Contingency Table, Null Hypothesis, Aspirin
Document Summary
The chi-square distribution and the analysis of frequencies (2) Example: in the physician"s health study 11,037 physicians took 325mg. Aspirin, and 11,034 took a placebo every other day over a 10-year period. 104 people in the aspirin group had a heart attack, and 189 people in the placebo group had a heart attack. We want to test h0: p1=p2, ha: p1 p2, where. P1 is the probability of a heart attack for the aspirin group, and p2 for the placebo group. 22,071 this is called a 2x2 contingency table in general, we want to use the data from an r c (rows x columns) contingency table to test: Ha: pi ps for at least one pair (i, j) This null hypothesis is called the homogeneity hypothesis step 1: calculate frequencies and ei are the frequencies expected when h0 is true step 2: reject h0 if x2(r-1)(c-1)>x2(r-1)(c-1),1- .