An efficient and flexible multiplicity adjustment for chi-square endpoints

Area of Science:

• Biostatistics
• Statistical Methodology
• High-Dimensional Data Analysis

Background:

• Multiplicity adjustment is crucial for controlling Type I errors in studies with multiple endpoints.
• High-dimensional data presents unique challenges for traditional statistical adjustments.
• Ensuring reproducibility requires robust error control in statistical testing.

Purpose of the Study:

• To propose a novel, fast, and efficient multiplicity adjustment method.
• To strictly control the Type I error rate for high-dimensional chi-square distributed endpoints.
• To offer a flexible method applicable to various correlation structures.

Main Methods:

• Development of a new multiplicity adjustment procedure.
• Application to families of high-dimensional chi-square distributed endpoints.
• Evaluation of computational feasibility and statistical power.

Main Results:

• The proposed method strictly controls the family-wise error rate.
• It demonstrates greater statistical power than the Bonferroni adjustment.
• It is computationally more feasible than existing methods in high-dimensional settings.

Conclusions:

• The new method provides a powerful and computationally efficient solution for multiplicity adjustment.
• It enhances the credulity and reproducibility of results in high-dimensional studies.
• The method is applicable to diverse correlation structures and large-scale testing scenarios.