Issues in Weighting the 2009 Residential Energy Consumption Survey

• 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): Rachel Harter and Mike Kwanisai*+ and Ashley Amaya
• Companies: NORC at the The University of Chicago and NORC at the The University of Chicago and NORC at the The University of Chicago
• Keywords: area probability ; weights ; multiplicity ; adjustment ; poststratification

Abstract

The 2009 Residential Energy Consumption Survey was unusual in that it required a massive expansion of the sample design while preserving the 2005 RECS design for future use. Many primary and secondary sampling units in the area probability design had two probabilities of selection, one from 2005 and another from 2009, but sufficient information to compute joint probabilities was not available. The probabilities and separate base weights were scaled and joined using a multiplicity adjustment to form consolidated base weights. Dummy housing unit records were introduced into the nonresponse adjustments to account for missing secondary sampling units. Defining cells for poststratification was a balance between sample size and variability in the ACS control totals.