NSHAP's Wave 2 Nonresponse Weight Adjustment with Some Wave 1 Nonresponses

• Activity Number: 87
• Type: Contributed
• Date/Time: Sunday, August 4, 2013 : 4:00 PM to 5:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Steven Pedlow*+ and Colm O'Muircheartaigh and Phil Schumm
• Companies: NORC at the University of Chicago and NORC at the University of Chicago and The University of Chicago
• Keywords: Weighting Cells ; Panel Surveys

Abstract

The National Social Life, Health, and Aging Project (NSHAP) is a longitudinal study of the health of older adults, concentrating on the role of social relationships in the aging process. In 2005 and 2006, NORC and a group of investigators at the University of Chicago interviewed a nationally representative sample of adults aged 57 to 85 for Wave I. In 2010 and 2011, these respondents as well as their spouses or cohabitating romantic partners were interviewed for Wave II.

Unlike most longitudinal surveys, we also attempted to interview in Wave II individuals who were sampled but declined to be interviewed in Wave I. This creates two alternatives in computing a non-response weight adjustment for Wave II. One possibility is that the Wave I non-respondents who responded in Wave II are most like the non-respondents to both waves, and thus should be in the same cells as the non-respondents to both waves, resulting in high weights for these cases and a larger design effect. An alternative is that other characteristics better explain the non-response, which divides the non-respondents to both waves across many cells, resulting in a smaller design effect. This paper evaluates and compares these two alternatives.