There are numerous ways to use the survey 1 dataset from the website as there are several categorical variables. Here are some examples for the tests described in this chapter:
a. Conduct a test of a single proportion: Is the sample proportion of females equal to .50?
b. Conduct a test of two independent proportions: Is there a difference between the sample proportion of females who are right-handed and the sample proportion of males who are right-handed?
c. Conduct a test of two dependent proportions: Is there a difference between the sample proportion of students' mothers who are right-handed and the sample proportion of students' fathers who are right-handed?
d. Conduct a chi-square goodness-of-fit test: Do the sample proportions for the political view categories differ from their expected proportions (very liberal = .10, liberal = .15, middle of the road = .50, conservative = .15, very conservative = .10)? Determine if the assumptions of the test are met. Determine and interpret the corresponding effect size.
e. Conduct a chi-square goodness-of-fit test to determine if there are similar proportions of respondents who can (vs. cannot) tell the difference between Pepsi and Coke. Determine if the assumptions of the test are met. Determine and interpret the corresponding effect size.
f. Conduct a chi-square test of association: Is there an association between political view and gender? Determine if the assumptions of the test are met. Determine and interpret the corresponding effect size.
g. Compute a chi-square test of association to examine the relationship between if a person smokes and their political view. Determine if the assumptions of the test are met. Determine and interpret the corresponding effect size.