Address-Based Sampling with a Field Component for the CICPE 2010

• Activity Number: 65
• Type: Topic Contributed
• Date/Time: Sunday, July 31, 2011 : 4:00 PM to 5:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Steven Pedlow*+
• Companies: NORC at the The University of Chicago
• Keywords: Census Integrated Communication Program Evaluation ; address based sampling

Abstract

The 2010 Census Integrated Communication Program Evaluation (CICPE 2010) included three waves of data collection to understand attitudes, knowledge, and intent/participation for the 2010 Decennial Census. Each wave consisted of an address-based sample (ABS) in which we attempted to match all selected addresses to a phone number. The primary mode of data collection was by phone, but we also attempted an in-person interview for a subset of unmatched addresses as well as a subset of those addresses where the matched phone number did not lead to a completed interview. The design motivation was to save costs versus a full in-person survey while replacing the disadvantages of phone surveys with the advantages of in-person surveys on coverage and response rates.

This talk will discuss various aspects of data collection, comparing the assumptions with our actual experience on phone number matching rates, eligibility rates, response rates, and the necessary subsampling rates to give lessons that we applied in later waves, and to provide guidance for other researchers that want to use address based sampling with or without a field component.