Adjusting for multiple testing when reporting research results: the Bonferroni vs Holm methods

Area of Science:

• Biostatistics
• Public Health Methodology

Background:

• Multiple testing adjustments are crucial in public health research for accurate results.
• The Bonferroni procedure is commonly used but can be overly conservative, reducing statistical power.
• Holm's procedure is statistically more powerful than Bonferroni but less frequently adopted.

Purpose of the Study:

• To present an accessible P-value adjustment method based on Holm's procedure.
• To encourage the wider adoption of Holm's procedure in public health research.
• To enhance statistical power in public health studies through improved multiple testing adjustments.

Main Methods:

• A P-value adjustment method derived from Holm's procedure is described.
• The method focuses on P-value manipulation for easier interpretation by researchers.
• Comparison with the Bonferroni procedure is implicitly suggested through the rationale for adoption.

Main Results:

• The proposed P-value adjustment facilitates the application of Holm's procedure.
• This method aims to maintain higher statistical power compared to Bonferroni.
• The adjustment is presented to overcome barriers to Holm's procedure adoption.

Conclusions:

• Holm's procedure, via P-value adjustment, offers a statistically superior approach to multiple testing in public health.
• Adoption of this method can lead to more robust and powerful research findings.
• Simplifying the application of Holm's procedure is key to its increased use in the field.