Commercial telephone samples and component outcome rates - ARCHIVED

This paper discusses in detail issues dealing with the technical aspects of designing and conducting surveys. It is intended for an audience of survey methodologists. Outcome rates in telephone surveys are usually based on an entire sample. However, telephone samples from commercial sample vendors contain identifiable subsets of records with very different probabilities of obtaining particular dispositions. In such a case, component outcome rates could vary in ways unrelated to rates based on the entire sample. The 2000 Behavioral Risk Factor Surveillance System (BRFSS) survey examines the degree to which selected outcome rates (by state, for different subsets of records) correlate with corresponding global rates. Although correlations tend to be large, not all cases are, making it worthwhile to examine component outcome rates.