A framework for sample size calculations in longitudinal surveys to measure net and gross changes

Area of Science:

• Survey Methodology
• Statistical Power Analysis
• Longitudinal Data Analysis

Background:

• Longitudinal surveys collect repeated measurements from the same units to assess individual (gross) change.
• Fresh samples are often added to measure aggregate (net) change.
• Existing methods lack comprehensive sample size determination for simultaneous net and gross change measurement.

Purpose of the Study:

• To propose a novel framework for calculating sample sizes in longitudinal surveys.
• To enable concurrent measurement of net and gross changes in means or proportions.
• To provide a tool for researchers to determine optimal panel and fresh sample sizes.

Main Methods:

• Developed a statistical framework for sample size calculation in longitudinal surveys.
• Formulated methods to compute panel and fresh sample sizes based on desired net and gross change levels.
• Created an R package (nchange) to implement the proposed framework.

Main Results:

• The proposed framework provides a method for sample size determination in longitudinal surveys measuring both net and gross changes.
• The nchange R package facilitates the application of this framework.
• Demonstrated the framework's utility in determining sample sizes for specific statistical power.

Conclusions:

• The developed framework and R package address a critical gap in longitudinal survey design.
• Researchers can now determine appropriate sample sizes for measuring both net and gross changes concurrently.
• This will enhance the statistical power and efficiency of longitudinal studies.