Power Analysis and Sample Size¶
Statistical power analysis and sample size estimation for clinical trials.
Features¶
- Power calculation for given sample size
- Sample size estimation for target power
- Effect size interpretation (Cohen's d)
- Multiple comparison considerations
Cohen's d Reference¶
| Effect Size | Cohen's d | Interpretation |
|---|---|---|
| Small | 0.2 | Subtle difference |
| Medium | 0.5 | Moderate difference |
| Large | 0.8 | Obvious difference |
Sample Size Guide (80% power, alpha=0.05)¶
| Effect Size | N per arm |
|---|---|
| Small (0.2) | ~394 |
| Medium (0.5) | ~64 |
| Large (0.8) | ~26 |
Features Demonstrated¶
- Analytical power calculation
- Sample size estimation
- Dropout adjustment
- Power curves
Files¶
| File | Description |
|---|---|
| python.py | Python implementation |
| julia.jl | Julia implementation |
Expected Output¶
Power Analysis
==============
Given: n=50 per arm, effect=0.5, alpha=0.05
Calculated power: 69.7%
Sample Size Estimation
======================
Target: 80% power, effect=0.5, alpha=0.05
Required N per arm: 64
Total N: 128
Achieved power: 80.1%
With 15% dropout adjustment:
Required N per arm: 76
Total N: 152