Effects of Sampling and Screening Strategies in an RDD Survey

The Monitoring Sexually Transmitted Infections Project is an RDD telephone survey being conducted in Baltimore, Maryland from 2006 through 2009. The project screens households for people 15-35 years of age and then randomly selects one eligible person. To gain efficiency and reduce costs, three methodological changes were adopted after several months of data collection. These changes were 1) in the wording of the screening questions aimed at identifying households with someone 15-35, 2) implementing a sampling strategy that incorporated the use of lists that identified the ages of residents within some households, and 3) altering the probability of selection of an eligible respondent within a household to favor selecting someone who initially answered the telephone. This paper will examine the effects of these strategies on survey costs and sample composition.