Question

A random sample of voters from three cities is asked to indicate political party affiliation. The results are given in the table.

City	Political Party Affiliation
	Democrat	Independent	Republican
Miami	211	57	132
Austin	126	84	190
Detroit	158	96	146

To test whether the proportion of party affiliations for each city is the same, which inference procedure is appropriate to use?

Answer 1

To test whether the proportion of party affiliations for each city is the same, the chi-square test for independence inference procedure is appropriate to use because the two variables political party affiliation and city are categorical variables. The categorical variable political party affiliation has three levels and the other categorical variable city has three level. So, for checking whether the proportion of party affiliations for each is the same or not, we need to use the chi-square test for independence between two categorical variables.