Using a Dual Frame Sampling Design to Increase the Efficiency of reaching Population subgroups in a Telephone Survey
Roe, D., & Currivan, D. (2004). Using a Dual Frame Sampling Design to Increase the Efficiency of reaching Population subgroups in a Telephone Survey. In American Association for Public Opinion Research 59th Annual Conference AAPOR.
The effort and cost required to reach households and complete interviews in random-digit dialing (RDD) telephone surveys has increased over the past several years. Effort and cost are even greater in RDD surveys when the sample design specifies screening the population for specific subgroups, such as age or ethnic groups. When the probability of reaching respondents in a specific subgroup is sufficiently low, the cost of using a standard RDD approach can be prohibitive. An alternative strategy is to supplement RDD numbers with numbers selected from directory listings. Using listed numbers can significantly improve the probability of reaching eligible respondents and thereby lower the effort and cost of screening households and completing interviews. Still, directory- listed sample frames have the important shortcoming of excluding the growing number of households that do not currently have listed numbers. Furthermore, additional information on the listed sample frame on how likely household members belong to a particular subgroup may be of limited accuracy. The potential result of these shortcomings of listed sample frames is bias in survey estimates. The goal of this research is to better understand the costs and benefits of supplementing an RDD sample with a listed sample in a national survey that targets respondents in certain age and ethnic groups. The analysis compares the two sampling frames on both key outcomes of the data collection effort and on substantive results among completed interviews. Key data collection outcomes include rates of working and residential numbers, eligibility rates, and completion rates. Among completed interviews, we examine differences in the targeted demographics (age and ethnicity), as well as other demographic characteristics and substantive indicators. This research will provide some evidence on the potential of dual- frame designs to provide accurate data on population subgroups with less effort and cost compared to RDD methods.
