Global goodness-of-fit tests for group testing regression models

Area of Science:

• Biostatistics
• Epidemiology
• Infectious Disease Modeling

Background:

• Group testing (pooled testing) is a cost-effective method for infectious disease screening, offering high precision compared to individual testing.
• Regression models have been developed for group testing data to incorporate individual covariate information.
• Existing statistical methods lack tools to assess the adequacy of these regression models.

Purpose of the Study:

• To introduce novel global goodness-of-fit tests for regression models applied to group testing data.
• To provide methods for checking the validity and adequacy of statistical models used in pooled testing analyses.
• To evaluate the performance of these new tests in various scenarios.

Main Methods:

• Development of global goodness-of-fit tests specifically designed for regression models with group testing data.
• Simulation studies to assess the small-sample size and power properties of the proposed tests.
• Evaluation across different pool composition strategies to understand their impact on test performance.

Main Results:

• The study presents effective goodness-of-fit tests for regression models in group testing.
• Simulation results demonstrate the performance characteristics (size and power) of these tests under various conditions.
• The methods are illustrated using real-world infectious disease data.

Conclusions:

• The developed goodness-of-fit tests provide essential tools for validating regression models used with group testing data.
• These methods enhance the reliability of statistical analyses in infectious disease screening and other biomedical applications.
• The application to HIV and infertility prevention data highlights the practical utility of the proposed tests.