Sampling Children for a Study of the State Children's Health Insurance Program (SCHIP)

• Activity Number: 209
• Type: Contributed
• Date/Time: Tuesday, August 13, 2002 : 10:30 AM to 12:20 PM
• Sponsor: Section on Survey Research Methods*
• Author(s): Mourad Touzani*+ and Frank Potter and Amy Zambrowski
• Affiliation(s): Mathematica Policy Research, Inc. and Mathematica Policy Research, Inc. and Mathematica Policy Research, Inc.
• Keywords: Children Survey ; Sampling Design ; SCHIP

Abstract

Surveys of children to study the effects of health programs pose complex sampling issues. In 1997, Congress established the SCHIP program, which allowed states to create a new program, to expand Medicaid, or both. Congress also mandated an evaluation focusing on ten states representing a mix of program designs and geographic regions and a high proportion of the nation's low-income uninsured children. The legislation mandated a survey which was designed as a cross-sectional, dual-sample telephone survey (a nonclustered sample and a clustered sample with field follow-up). Respondents in a state include 600 new and 600 established enrollees and 600 recent disenrollees, totaling 18,000. Because data on past events are collected on new enrollees and recent disenrollees, sampling is conducted two or three times within six months for each state to reduce recall effects (i.e., the time between state frame development and data collection). We will discuss the sampling design and the sampling frame issues and solutions we developed. The issues involve the access and use of state data files for sampling frame development, the use of a dual sample, and sampling at repeated close time intervals.