Friedman Test
=
Tap a term to see what it means, or .
Calculate
Example: six tasters scored three wines. Each taster has their own yardstick — some score high, others low —, so what matters is the order within each one. Does any wine tend to come out ahead? Change the values to work out your own case.
Step by step
What each term means
- Friedman Statistic (χ²_F)
- Measures how far the treatments' rank sums are from what is expected. With no difference, it roughly follows a chi-square with k − 1 degrees of freedom.
- Scale Factor
- n is the number of blocks and k the number of treatments.
- Sum of Squared Ranks (Rⱼ²)
- Rⱼ is treatment j's rank sum, with ranks given within each block, 1 to k. A treatment always ahead builds up high ranks.
- Expected Value (3n(k + 1))
- What the sum would give if every treatment had the same mean rank, (k + 1)/2.
When to use it, assumptions and alternatives: the Friedman test in the statistical tests catalog.