Response Rates in Random Digit Dialing RDD Telephone Surveys

A. INTRODUCTION Over the past 20 years there has been a dramatic increase in the use of random digit dialing (RDD) telephone surveys by both private and public organizations to collect information on a variety of topics. The term "random digit dialing" refers to the use of all possible telephone numbers as a sampling flame for telephone surveys. A review of RDD and other telephone survey sampling methods are discussed by Lepkowski (1988). As the proportion of the United States household population without telephones decreased to under 10 percent in the 1970s, national RDD surveys became more feasible. As the concem over telephone undercoverage decreased, more attention began to be focused on response rates and the quality of data obtained in RDD surveys. Response rates over 70 percent were being obtained by many organizations in the 1980s with a reasonable amount of effort. Several government organizations adopted a RDD methodology for their major household surveys because of lower cost and other attractive features of the RDD methodology. The National Center for Education Statistics has used a RDD methodology for their National Household Education Survey (NHES) since the late 1980s. Response rates in the 1991 and 1993 NHES exceeded 70 percent for almost all components of the surveys. In 1995 and 1996, however, the NCES response rate declined by about 10 percentage points for most components (NCES technical report 97-948). The objectives of this project were to: identify RDD surveys sponsored by government and other organizations over the past 5 years and document the response rates obtained in the surveys; review and summarize methods used by government agencies and survey research organizations to calculate and report RDD response rates; to the extent possible, identify the key correlates of response rates in RDD surveys; and review and summarize standards and guidelines used by government agencies and survey research organizations for RDD survey response rates.