Improving mail survey response rates in Japan: empirical tests for envelopes, request letters, questionnaires, and schedules

While many studies examine tools and techniques to improve survey response rates, it remains unclear as to which are effective, and response rates continue to decline. In this study, we examine the effect of four variables expected to influence mail survey response rates: (1) the use of semi-transparent envelopes versus ordinary paper envelopes for delivery, (2) the level of content in the request letter (simple or detailed), (3) the length of the questionnaire, and (4) the survey schedule. We found that the total response rate was 45.4%, while those of the experimental conditions were as follows: (1) semi-transparent versus ordinary paper envelopes, 46.2% versus 44.7%; (2) simple versus detailed request letters, 44.7% versus 46.2%; (3) 4-page versus 8-page questionnaires, 46.0%versus 44.8%; and (4) holiday versus ordinary day, 45.6% versus 45.3%.We found no significant differences in response rates for any of the experimental variables; that is, the logistic regression analysis revealed no relationship between the four experimental conditions and the response rate.

Automated summary

The study examined four variables that were expected to influence mail survey response rates but found no significant differences in response rates for any of the experimental variables.