An alternative measure of response rate in random-digit-dialing surveys that screen for eligible subpopulations

Abstract

The evaluation of the level of nonresponse in a survey and the impact of that nonresponse on the survey estimates is an important issue in survey research. It is important to apply the most accurate method of calculating response rates in order to understand the components of nonresponse and to be in a better position to incorporate modifications in the survey procedures as necessary and feasible. Response rates have been defined in a number of ways for RDD surveys. The standard/traditional method of calculating response rates involves the product of three rates -- the resolution rate, the screening completion rate, and the interview completion rate. The coverage rate (population coverage) has often been reported separately from the response rate. The method presented in this paper has the advantage of combining the coverage rate and the response rate in the calculation of the overall response rate.