41 – Nonresponse Adjustment - 1
Noncoverage Adjustments in a Single-Frame Cell-Phone Survey: Weighting Approach to Adjust for Phoneless and Landline-Only Households
Nadarajasundaram Ganesh
NORC at the University of Chicago
Meena Khare
National Center for Health Statistics, CDC
Elizabeth Ormson
NORC at the University of Chicago
Wei Zeng
NORC at the University of Chicago
Jenny Jeyarajah
National Center for Immunization and Respiratory Diseases
David Yankey
National Center for Immunization and Respiratory Diseases
Zhen Zhao
National Center for Immunization and Respiratory Diseases
Kirk M. Wolter
NORC at the University of Chicago,
In 2011, the National Immunization Survey (NIS) began using a dual-frame landline and cell-phone sample design to monitor vaccination coverage rates among children 19-35 months. As of 2012, coverage of the cell-phone sampling frame has increased to 92.9% of all age-eligible children in the NIS. Thus, a single-frame cell-phone sample design could be a viable option for the NIS, but such a sample design would not cover children living in landline-only and phoneless households. In order to adjust for the noncoverage of children living in landline-only and phoneless households, we propose two distinct adjustment methods. The first option is to identify a subset of cell-phone households with sampled children who have similar socio-demographic characteristics as those from landline-only and phoneless households and use a weighting class based approach to adjust for the noncoverage. The second option is to adjust for the noncoverage by using socio-demographic characteristics in a raking adjustment step. We implemented the proposed methods with 2012 NIS data and compared survey estimates generated under a single-frame cell-phone sample design with similar estimates from the current dual-frame sample design.