Sample Size Comparison

Arocena Thomas, Chaltiel Dan & Lourdessamy Anne

Welcome to the website dedicated to presenting the results of the Sample Size Comparison project. This site brings together several complementary sections designed to guide you through the work:

• Book — the core of the project, where each study design is documented with its methodology, assumptions, and comparative results.
• Interactive — dynamic visualisations that allow you to explore the results in more detail, adjust parameters, and better understand how sample size choices influence performance.
• More — additional material related to the main topic, including slides and side analyses.
• Code — a transparent view into the full computational workflow, from simulation engines to plotting functions, ensuring reproducibility and openness.

Together, these sections provide a unified space to navigate the project, understand the methodological insights, and interact directly with the underlying analyses.

Results summary

Below is a concise overview of the main findings from the Sample Size Comparison project.
This summary is intended for readers who want the key takeaways without exploring the full “Book” section.

The table compares, for each endpoint and design type, the statistical test used, and how well the implemented R calculations matched the software(East or nQuery) reference values.

Endpoints	Design type	Test	R/Software matching
survival	2‑Arm fixed design	Log‑rank	🟢 Perfect
survival	2‑Arm group‑sequential design	Log‑rank	🟡 Good
survival	1‑Arm fixed design	Log‑rank	🔵 Poor
Binary	2‑Arm fixed design	Exact	🟡 Good
Binary	2‑Arm fixed design	Z‑Pooled	🟢 Perfect
Binary	2‑Arm group‑sequential design	Z‑Pooled	🟢 Perfect
Binary	1‑Arm fixed design	Exact	🟡 Good
Binary	1‑Arm fixed design	1‑Arm Z test	🟢 Perfect

Additional content: