Question

Suppose that a magazine predicted that Candidate A would defeat Candidate B in a certain election. They conducted a poll of home owners with a response rate of 25​%. On the basis of the​ results, the magazine predicted that Candidate A would win with​ 57% of the popular vote.​ However, Candidate B won the election with about​ 62% of the popular vote. At the time of this​ poll, most home owners belonged to the party of Candidate A. Name two biases that led to this incorrect prediction. Choose the correct answer below. A. Sampling​ bias: Using an incorrect frame led to undercoverage. Response​ bias: The way the poll was administered showed bias. B. Nonresponse​ bias: The low response rate caused bias. Response​ bias: The way the poll was administered showed bias. C. Sampling​ bias: Using an incorrect frame led to undercoverage. Nonresponse​ bias: The low response rate caused bias.

Answer 1

The correct answer is: C. Sampling bias: Using an incorrect frame led to undercoverage. Nonresponse bias: The low response rate caused bias.

The result of the survey was erroneous because it had two very marked biases: the sampling bias and the non-response bias.

The sampling bias is clearly seen because the surveys were conducted only on the homeowners, who were from the party of candidate A and therefore had a preference for candidate A.

The non-response bias is evidenced because the response rate was very low (only 25%).

We have no information about how the surveys were conducted or the cause of the low response. The magazine did not elaborate on this topic nor did it propose any inference. In consequence we can't asseverate that there was a response bias in the result of the survey.