Sample size estimation for comparing two or more treatment groups in clinical trials.

Area of Science:

• Statistical methodology
• Biostatistics
• Research design

Background:

• Existing methods for sample size estimation often rely on complex formulas and tables.
• Current techniques are typically limited to comparing only two groups.
• Determining sample sizes for more than two groups is significantly more complicated.

Purpose of the Study:

• To introduce a simplified method for sample size estimation.
• To address the limitations of existing methods for comparing multiple groups.
• To provide a practical tool for researchers across different study designs.

Main Methods:

• A simple linear nomogram was developed for sample size calculation.
• The nomogram's application was demonstrated using practical examples.
• The method accommodates parallel group, ordered parallel group, and factorial designs.

Main Results:

• The proposed linear nomogram offers a straightforward approach to sample size estimation.
• It simplifies calculations for comparing two population means.
• The nomogram effectively extends to scenarios involving more than two groups and complex designs.

Conclusions:

• The linear nomogram provides an accessible and efficient solution for sample size determination.
• This tool can be widely applied in statistical research, particularly in biostatistics.
• It facilitates more straightforward sample size planning for diverse experimental and observational studies.