This calculator performs the Nemenyi post-hoc test, typically used after a significant Friedman test for related samples (blocked designs).
Each row in your input data (manual or CSV) is treated as a "block" or "dataset", and each column is a "group" or "algorithm".
The calculator will rank data *within each block*.
For more accurate q-values for degrees of freedom not listed below, please refer to a Studentized Range Q table.
Upload Data from CSV
Select a CSV file. The first row is assumed to be group headers (e.g., algorithm names), and the first column is assumed to be block/dataset names.
Results
Average Ranks and Group Rank Sums
Algorithm
Average Rank
Group Rank Sum
These are the average rank and total sum of ranks for each group (algorithm) after ranking has been performed independently within each row (dataset).
Pairwise Comparisons
Comparison
Absolute Difference in Rank Sums
Critical Difference (CD)
Significance
*Pairs where the absolute difference in rank sums is greater than the Critical Difference (CD) are considered statistically significant at the chosen $\alpha$ level.
Statistically Significant Differences
Comparison
Absolute Difference in Rank Sums
Critical Difference (CD)
No statistically significant differences were found between any pairs of algorithms at the chosen $\alpha$ level.