Response rate and non-response bias analysis
During the months of January-March, 2010, 20,000 emails were sent randomly to
respondents in the US who have an email address through rent-a-list.com’s database.
Only 1,807 completed the survey yielding a gross response rate of 9 percent. At this
stage, all responses that contained missing values were deleted from the study. Further
filtering of the survey responses based on the first qualifying question, we identified
899 respondents who had stayed in a hotel within the last 12 months. The second
qualifying question eliminated 325 respondents (36.2 percent) who had not used their
credit cards to pay for any of the services in a hotel. Based on that, 574 fully completed
responses were qualified for the study and hence, were used for further analysis. The
study yielded 2.8 percent net response rate. A non-response analysis using wave
analysis (early versus later respondents) was conducted (Rylander et al., 1995) and
revealed no significant difference between the two groups. Based on this it was
concluded that the sample did not suffer from non-response bias.