Nonparametric Inference

One Sample Wilcoxon test

one_wilcoxon_inference(wt ~ 1, data = mtcars2)

response	n	pseudomedian	conf.low	conf.high	null	V	p.value
wt	32	3.19	2.83	3.51	0.000	528	< 0.0001
Wilcoxon signed rank test with continuity correction (two.sided), with 95% confidence intervals.
Approximate p-value used, due to ties.
Approximate confidence interval used, due to ties.

Separately by another categorical variable

one_wilcoxon_inference(wt ~ am, data = mtcars2)

response	variable	n	pseudomedian	conf.low	conf.high	null	V	p.value	footnote
wt	am = automatic	19	3.63	3.44	4.28	0.000	190	0.0001	1,2,3
wt	am = manual	13	2.39	2.02	2.78	0.000	91.0	0.0002	4
1 Wilcoxon signed rank test with continuity correction (two.sided), with 95% confidence intervals.
2 Approximate p-value used, due to ties.
3 Approximate confidence interval used, due to ties.
4 Wilcoxon signed rank exact test (two.sided), with 95% confidence intervals.

Two Sample Wilcoxon test

two_wilcoxon_inference(wt ~ am, data = mtcars2)

response	variable	pseudomedian	conf.low	conf.high	null	W	p.value
wt	am: automatic - manual	1.28	0.79	1.82	0.000	230	< 0.0001
Wilcoxon rank sum test with continuity correction (two.sided), with 95% confidence intervals.
Approximate p-value used, due to ties.
Approximate confidence interval used, due to ties.

Kruskal-Wallis test

kruskal_wallis_test(wt ~ am, data=mtcars2)

response	variable	df	chisq	p.value
wt	am	1	16.9	< 0.0001
Kruskal-Wallis rank sum test

Pairwise Wilcoxon tests

pairwise_wilcoxon_inference(wt ~ cyl, data = mtcars2)

response	variable	pseudomedian	conf.low	conf.high	null	W	p.value	p.adjust
wt	cyl: 4 - 6	−0.93	−1.52	−0.16	0.000	8.00	0.0066	0.020
wt	cyl: 4 - 8	−1.62	−2.27	−0.97	0.000	1.00	< 0.0001	0.0001
wt	cyl: 6 - 8	−0.66	−1.90	−0.13	0.000	9.00	0.0032	0.0095
Wilcoxon rank sum test with continuity correction (two.sided), with 95% confidence intervals, adjusted for 3 comparisons using the Bonferroni method.
Approximate p-value used, due to ties.
Approximate confidence interval used, due to ties.
p-values adjusted for 3 multiple comparisons using the Bonferroni method.

Paired Wilcoxon test

paired_wilcoxon_inference(score2 - score1 ~ 1, data = passfail)

response	pseudomedian	conf.low	conf.high	null	V	p.value
score2 - score1	3.1	0.5	5.5	0.000	881	0.019
Wilcoxon signed rank test with continuity correction (two.sided), with 95% confidence intervals.