Wilcoxon signed-rank test — setup.

medium

Answer

$Non-parametric alternative to the paired t-test.$
$d_{i}$
$Null: distribution of differences is symmetric around zero.$
$Robust to non-normal differences, sensitive to outliers still (unlike sign test).$
$Uses ranks so drops information but keeps power when normality is doubtful.$

Check yourself — multiple choice

Independent samples
Non-parametric paired test: sign + rank $\mathrm{of}\; \mid d_{i}$ |; null = differences symmetric around 0; more robust than paired t under non-normality
Requires normality
Same as Mann-Whitney

Wilcoxon signed-rank: non-param paired test on $\mathrm{ranked}\; \mid d_{i}$ |.

#hypothesis-testing#non-parametric

215 interview questions in this topic.