Abstract
n past government surveys, incentive payments have proven successful in increasing response rates when surveying hard-to-interview populations. In order to examine the impact of paying respondents in its own client follow-up survey, the Alcohol and Drug Services Study included an incentive substudy. The sample for this substudy, made up of outpatient non-methadone clients from the ADSS' main study (respondents paid $25) and three incentive cohorts (respondents paid $0,$10, or $35), was specifically created to determine the impact of varying incentives on such methodological issues as response rate, sample composition, and response quality. In the current analysis we ask whether payment of incentives can also be cost effective. Specifically, we ask whether incentives can actually increase response rate to the extent that average cost per respondent drops? To answer this question we associate cost information from the ADSS data collection with corresponding survey completion statistics to determine the impact of payment level and other important factors in modeling cost per survey respondent.