Statistical tests
Shapiro–Wilk test
Implemented
statistics.shapiro_wilkDefinition
Tests the null hypothesis that a sample comes from a normal distribution.
Formula
W = (Σ aᵢ x₍ᵢ₎)² / Σ (xᵢ − x̄)²
Inputs and outputs
- sample
Returns: StatisticalTestResult
Assumptions
No assumptions beyond valid, aligned inputs of the documented types.
Limitations
- With large samples, trivial departures from normality become significant.
Python API
import evalsuite as es
es.shapiro_wilk_test(scores_a)References
- Shapiro, S. S., & Wilk, M. B. (1965). An analysis of variance test for normality (complete samples). Biometrika, 52(3–4), 591–611.