How can you estimate power for an unusual test when no analytic formula is available?
StatisticsLevel: IntermediatePermutation experiment design
How can you estimate power for an unusual test when no analytic formula is available?
Hints
- Simulate under the alternative, not the null.
- A rejection indicator is Bernoulli.
Answer
Simulate independent datasets under a specified alternative, run the planned test, and estimate the rejection fraction. Report Monte Carlo uncertainty and the assumed effect/data model.
Source and licence: SciPy — resampling and Monte Carlo inference · BSD-3-Clause · answer: Quantoros editorial