Hierarchical Archimedean copula models for the analysis of binary familial data

Area of Science:

• Statistics
• Biostatistics
• Statistical Modeling

Background:

• Archimedean copulas are widely used for their simplicity and analytical properties.
• Their inherent exchangeable dependence structure limits applications in familial data with complex associations.
• Modeling binary familial data presents unique challenges due to dependence parameter constraints.

Purpose of the Study:

• To propose hierarchical Archimedean copulas for modeling familial data.
• To address the specific challenges of modeling binary familial data.
• To provide a flexible statistical tool for analyzing hierarchical dependence structures.

Main Methods:

• Development of hierarchical Archimedean copula models.
• Inference based on maximum likelihood estimation.
• Application to binary familial data analysis.

Main Results:

• Demonstrated the flexibility of hierarchical Archimedean copulas.
• Provided a framework for modeling complex familial dependencies.
• Successfully applied the method to binary familial data.

Conclusions:

• Hierarchical Archimedean copulas offer a powerful solution for familial data analysis.
• The proposed method effectively models hierarchical dependence structures, particularly for binary outcomes.
• This approach enhances statistical modeling capabilities for complex family-based studies.