How big should a clinical trial be?: a maths EPQ idea
A title to start from
How many patients does a clinical trial need for enough statistical power, and what goes wrong when it is too small?
Why it works as an EPQ
Hypothesis testing and power, with simulations showing what small trials do to conclusions.
Scope and difficulty
Solid. Solid. One outcome, two groups, simulation.
The maths
Builds on these A Level topics: Hypothesis testing · Statistical distributions · Sampling and data.
You would learn:
- Statistical power
- Effect size
- Sample-size formulas
One possible plan
- Explain type I and II errors in a trial.
- Derive a sample-size formula for comparing proportions.
- Simulate trials of different sizes.
- Judge the ethics of trials that are too small or too large.
Pitfalls
- Confusing statistical and clinical significance.
- Using trial results without reading the paper.
Where to start reading
- The Art of Statistics (David Spiegelhalter)
- Search for: sample size calculation two proportions power
Frequently asked questions
Is this a good maths EPQ idea?
It can be. We rate it Solid: it needs some new maths learned on your own. One outcome, two groups, simulation.
What maths does this EPQ use?
It builds on these A Level topics: Hypothesis testing; Statistical distributions; Sampling and data. You would learn: Statistical power; Effect size; Sample-size formulas.
Next step
Turn this idea into your own title, then check it in two minutes.
Check your titleSimilar ideas
- Is BMI a good measure?Is body mass index a sound measure, mathematically speaking?
- DNA evidence and the prosecutor's fallacyHow should a court interpret the chance of a DNA match, and how has probability been misused in trials?
- Models of tumour growthWhich growth model best describes tumour growth: exponential, logistic or Gompertz?
- Why giants cannot existWhy can't giants exist? Scaling laws in biology from ants to elephants
All maths in biology and medicine ideas · all 93 ideas
Your EPQ must be your own work. How to use help and AI honestly.