To improve response rates in telephone surveys, interviewers are often instructed by the polling company to state, "I am not selling anything" at the outset of the call. The effectiveness of the "I am not selling anything" strategy was investigated in the International Journal of Public Opinion Research (Winter 2004). The sample consisted of 29 different telephone surveys. However, in each survey about half the people were contacted by interviewers using the "I am not selling anything" introduction and the other half were contacted by interviewers using the standard (no mention of "not selling") introduction. Thus, for each of the 29 surveys, both the "not selling" and standard interviewing techniques were employed. Summary statistics on response rates (proportion of people called who actually responded to the survey questions) are given in the accompanying table. The goal of the researchers was to compare the mean response rates of the two interviewing methods with the specific purpose to determine if the mean response rate for "not selling" was higher than that for the standard.

a. Explain why the data should be analyzed as a paired difference experiment.
b. Analyze the data in the table using the independent samples t-test. Do you detect a significant difference between the mean response rates of the two methods using α = .05?
c. The researchers applied the paired difference t procedure and obtained an observed significance level of p-value = .001. Interpret this result if a = 0.5.
d. Compare the inferences you made in parts b and c.

Transcribed Image Text:

"Not selling Standard Introduction Introduction Number of Surveys Mean Response Rate Standard Deviation 29 262 .12 29 246