I'm learning statistics and while playing with post hoc tests, I noticed that Dunn's test returns the same p-value twice. I tried a different data set and the error repeated itself.
import pandas as pd
from scipy.stats import kruskal
from scikit_posthocs import posthoc_dunn
data = {
'Group': ['A', 'A', 'A', 'B', 'B', 'B', 'C', 'C', 'C'],
'Values': [10, 15, 12, 18, 20, 22, 25, 28, 30]
}
df = pd.DataFrame(data)
statistic, p_value = kruskal(df[df['Group'] == 'A']['Values'],
df[df['Group'] == 'B']['Values'],
df[df['Group'] == 'C']['Values'])
print(f"Kruskal-Wallis p-value: {p_value}")
posthoc_results = posthoc_dunn(df, val_col='Values', group_col='Group', p_adjust='holm')
print("\nDunna's Test Results:")
print(posthoc_results)
And i get. Dunna's Test Results:
A B C
A 1.000000 0.359425 0.021871
B 0.359425 1.000000 0.359425
C 0.021871 0.359425 1.000000
why the p-value in the comparison of A to B is the same as B to C