'Unmatched' RDD and ABS Sample: Are We Reaching the Same Population?

• Activity Number: 654
• Type: Topic Contributed
• Date/Time: Thursday, August 5, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Anna Fleeman and Robin Gentry*+ and Nicole Wasikowski
• Companies: Arbitron Inc. and Arbitron Inc. and Arbitron Inc.
• Keywords: Address Based Sampling ; Unmatched Phone ; Sampling

Abstract

Arbitron uses a combination of RDD landline phone sample and address-based sample (ABS) to recruit households for its radio ratings service. For RDD sample we attempt to match addresses, whereas for the ABS we attempt to match phone numbers. In theory, the sample for which neither address nor phone matches are available come from the same population (the same commercial matching databases are used). To test this hypothesis, the unmatched households in both frames - i.e., no-address RDD and no-phone ABS - were invited to participate in our diary-based radio ratings. Demographic characteristics, survey performance, and media habits of both groups give insight as to whether these two sampling approaches are reaching the same population. Given the increasing noncoverage of RDD landline frames and the growing interest in ABS, these findings are key to sound research and survey sampling.