An Assessment of the Effect of Misreporting of Phone Line Information on Key Weighted Estimates from Four National RDD Surveys

Ashley Bowers and Jeffrey M. Gonzalez


Random-digit dialing (RDD) surveys typically incorporate a weighting factor that corrects for unequal probabilities of selection due to multiple eligible landlines per household. The information that is used for the phone line adjustment is self-reported in one or two questions at the end of the survey and the wording varies substantially across studies. Few studies have addressed whether these phone line estimates are subject to measurement error that might affect the weighting adjustment. We compare estimates from several ongoing RDD surveys to those based on extended series of phone lines items from five studies that are viewed as having higher quality estimates. We also develop a set of modified weights that account for misreporting of phone lines in four RDD surveys and then re-estimate key figures from each survey to determine if the measurement error impacts survey estimates.