A dual frame design for sampling elderly minorities and persons with disabilities

Area of Science:

• Survey methodology
• Statistical sampling
• Public health research

Background:

• Multiple data sources offer potential sampling frames for population surveys.
• Multiple frame sample designs can be more advantageous than single frame designs in certain situations.
• However, multiple frames introduce complexities in variance, bias, sampling, data collection, and logistics.

Purpose of the Study:

• To address the implications of using multiple sampling frames in survey design.
• To investigate a proposed dual frame sampling approach for the National Health Interview Survey (NHIS).
• To evaluate sampling efficiencies and operational issues of supplementing the NHIS area frame.

Main Methods:

• Proposed a dual frame sampling approach for the National Health Interview Survey.
• Investigated supplementing the existing NHIS area frame sample.
• Utilized Social Security Administration files to select samples of specific populations.

Main Results:

• Presented findings on sampling efficiencies and operational considerations.
• Evaluated the feasibility of supplementing the NHIS area frame with targeted samples.
• Identified key issues related to variance, bias, and data collection in a dual frame design.

Conclusions:

• Dual frame sampling offers potential advantages but requires careful consideration of statistical and operational factors.
• The investigation provides insights into supplementing the NHIS with data from alternative frames.
• Further research is needed to optimize dual frame designs for comprehensive population surveys.