Exploratory Mediation Analysis of Dichotomous Outcomes via Regularization

Area of Science:

• Statistics
• Psychology
• Biostatistics

Background:

• Exploratory mediation analysis via regularization (XMed) is a novel technique for identifying mediators.
• Current XMed implementation is limited to continuous outcomes.
• Mediation analysis is crucial for understanding complex relationships in various scientific fields.

Purpose of the Study:

• To extend the exploratory mediation analysis via regularization (XMed) method to accommodate dichotomous outcomes.
• To enable XMed's application to both dichotomous mediators and dependent variables.
• To evaluate the performance of the extended XMed method.

Main Methods:

• The study extended the exploratory mediation analysis via regularization (XMed) technique.
• The extended method was adapted for dichotomous outcomes, including mediators and dependent variables.
• Simulations were conducted to compare the extended XMed with conventional mediation analysis methods.

Main Results:

• The extended XMed method demonstrated comparable sensitivity to traditional mediation analysis techniques (Sobel test, bootstrap methods).
• The extended XMed generally required approximately half the sample size compared to conventional methods.
• The approach was successfully demonstrated using an example of youth behavioral/emotional problems and alcohol use.

Conclusions:

• The extended XMed provides a powerful and more sample-efficient tool for mediation analysis with dichotomous outcomes.
• This advancement broadens the applicability of XMed in psychological, epidemiological, and clinical research.
• The findings suggest XMed is a valuable alternative for researchers dealing with dichotomous variables in mediation studies.