Question

A candidate running for office has asked for your assistance in analyzing data from a survey he has done. He surveyed people in his district and asked two questions – “Did you vote in the last election?” and “What was your salary last year?” He believes that people who have become disinterested in politics and thus no longer vote are the people who are having financial struggles. He plans to target non-voters in his campaign messages to get them back to the polls by promoting programs that will help them financially. Does the data support his theory? That is, is there evidence that the average salary for people who did not vote in the last election is lower than the average salary for people who did vote in the last election?

Which t-test is appropriate for this data?

Answer 1

For the given scenario, a two-sample t-test for difference between two population means or independent samples t-test for population means is an appropriate test for the given data. We have to test or check the researchers claim whether the average salary for people who did not vote in the last election is lower than the average salary for people who did vote in the last election.

Null hypothesis: H₀: The average salary for people who did not vote in the last election is same as the average salary for people who did vote in the last election.

Alternative hypothesis: H_a: The average salary for people who did not vote in the last election is lower than the average salary for people who did vote in the last election.