Basic statistical tests - parametric versus non-parametric tests

Definition

Parametric tests (e.g. t-test, ANOVA) assume the underlying data follows a specific distribution (typically normal) and are generally more statistically powerful when this assumption holds; non-parametric tests (e.g. Mann-Whitney U, Kruskal-Wallis) make no such distributional assumption, are more robust to outliers and skewed data, but are generally somewhat less powerful when the parametric assumption would have genuinely held
The choice between these two test families is a foundational methodological decision in essentially any quantitative clinical research analysis

4 more sections, plus exam facts

Premium unlocks every note across every specialty, and the full exam fact library behind it.