Mood's Median Test

Mood's Median Test — compare the MEDIANS of two or more groups, the robust way. Instead of averages, it counts how many values in each group land above or below the overall median and tests whether those splits differ. Skewed data and wild outliers barely move it. Get the chi-square, the p-value and clear above/below charts. Fully offline.

You have several groups of numbers — say house prices on three streets — and you want to know whether their TYPICAL value really differs. But prices are skewed and one mansion can drag an average around. Mood's median test sidesteps that: it asks only which side of the overall median each value falls on, then checks whether some groups sit high and others low more often than chance would allow.

No sign-up. No internet needed. No data collected. Just clear, instant statistics.

WHAT YOU CAN DO

• Test — type each group's values into a clean editor (add or remove groups and values with a tap) and instantly get the grand median, the above/below counts, the chi-square statistic and the p-value.

• Chart — see each group's share above (mint) vs at/below (coral) the grand median as split bars, a lane plot of where every value sits, and the chi-square curve with the observed value marked.

• Concept — learn the four steps in plain language — find the grand median, count above vs below, compare to what's expected, read the p-value — with offline, beautifully typeset formulas that never overlap.

• Simulate — slide three groups' medians apart with + / - steppered sliders, flip on a wild outlier, and watch the chi-square climb and the median test stay calm where a mean-based test would wobble.

• Examples — load real multi-group studies (skewed prices, a group with an outlier, three near-identical groups) with one tap.

WHAT YOU GET

• The grand median and each group's above/below counts.
• The chi-square statistic with its degrees of freedom (k - 1).
• The p-value and a plain-language verdict (medians differ / could be the same).
• Split bars, lane plots and a chi-square curve that make the result visual.

WHY YOU'LL LIKE IT

• Robust — built on above/below counts, so outliers and skew barely budge it.
• Tests the MEDIAN — the typical value, not an average a mansion can distort.
• Fully offline — every count, chart and formula works with no connection.
• Hands-on — type exact data, or build a sample with sliders and + / - steppers.
• Clean "Violet & Mint" design, easy on the eyes.
• Privacy-friendly — your data stays on your device.

Mood's median test is the simple, sturdy cousin of Kruskal-Wallis: it uses less of the data (just above/below), which makes it harder to fool with heavy skew. Reach for it when you want a quick, outlier-proof check of whether several groups share the same typical value.

Download it, enter your groups, and see whose median really stands apart.

Recent changes:
improve app performance and stability