Assessing and Adjusting for Response Error Using a Multi-Phase Survey Approach in the Residential Energy Consumption Survey

• Activity Number: 577
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Marilyn Worthy*+ and Katie Joseph
• Companies: Energy Information Administration and Energy Information Administration
• Keywords: multiphase survey ; nonresponse bias ; response error ; household survey ; RECS

Abstract

The Residential Energy Consumption Survey (RECS) is a quadrennial survey which collects energy data on a sample of U.S. households. The RECS questionnaire is technical, challenging respondents' capacity to report information on energy use, equipment, and related expenses. Householders who do not pay their energy bills directly, roughly 10% of the population, are more prone to sources of response error: difficulty understanding an energy question, not having access to information, and inability to produce an accurate response. EIA fills this quality gap by conducting a supplemental rental agent survey (RAS) for these households. We examine 2009 RECS data in households with both a household and a RAS interview to show patterns of response and nonresponse bias. We discuss the impact on key estimates and energy models, the utility of this approach in terms of the survey cost-error tradeoff,