Average Calculator

The mean is the everyday average: sum ÷ count. The median is the middle value once the numbers are sorted — with an even count, the average of the two middle ones. The mode is the value that appears most often. For 10, 20, 20, 30, 40 the sum is 120 and the count is 5, so the mean is 24. Sorted, the third of five values is the middle, so the median is 20. And 20 is the only value appearing twice, so the mode is 20 as well. All three are called "the average" in ordinary speech, which is exactly why disagreements about averages are so common — two people can quote different numbers from the same data and both be right. It is worth knowing what each one actually is. The mean is the balance point: if the values sat on a seesaw, it is where the pivot goes, because the deviations above and below it cancel to exactly zero. The median is the halfway marker: half the values fall on each side of it, regardless of how far away they are. The mode is simply the most popular value, and it is the only one of the three that must be a value that actually occurs in your data.

The mean uses every value, which is a strength and a weakness. Because a single extreme number enters the sum at full weight, outliers drag the mean toward them while the median barely moves. Take nine employees earning $50,000 and an owner earning $1,000,000. The mean pay is $145,000 — a figure nobody in the building receives, and one that would badly misrepresent the company in a job listing. The median is $50,000, which describes the typical employee exactly. This is why income, home prices and response times are almost always reported as medians. Those distributions are right-skewed: bounded below by zero, but with a long tail stretching upward. In a right-skewed distribution the mean sits above the median, and the gap between the two is itself a useful diagnostic — when mean and median are close, the data is roughly symmetric; when they diverge, something is pulling on the tail. A practical habit: report both. If they agree, either will do. If they disagree, the disagreement is the finding, and quoting only the flattering one is how honest numbers become misleading.

The mode gets the least attention of the three, mostly because it is nearly useless on continuous measurements. In a list of precise weights or timings, every value tends to appear exactly once, and there is no mode at all — which is what the calculator above reports as a dash. Where the mode is indispensable is categorical data, where the mean and median are not even definable. The most common shoe size sold, the most frequent error code, the most popular delivery day: you cannot average a shoe size across categories in any meaningful way, but you can certainly count which one comes up most. A distribution can also have more than one mode, and that is often the most interesting thing about it. Two clear peaks in a set of measurements usually means two different populations have been mixed together — two machines with different calibrations, two customer segments, two shifts. A mean computed across a bimodal set lands in the valley between the peaks and describes neither group. That is the real argument for looking at all four numbers this calculator returns rather than just the first one. A single summary statistic cannot tell you it is hiding a second peak.

The range is the simplest spread measure there is: max − min. For 10, 20, 20, 30, 40 the range is 30. Its weakness is structural. The range depends on exactly two numbers — the two most extreme ones — and ignores everything in between. That makes it the most outlier-sensitive statistic in common use, and it can only ever grow as you collect more data, since a bigger sample has more opportunity to find a new extreme. Ranges from samples of different sizes are not comparable. Two datasets can share a range of 30 while looking nothing alike: one with values evenly spread across the interval, another with everything clustered at 20 and a single stray at 40. For a real measure of spread you want something that uses all the values. The standard deviation reports the typical distance from the mean and pairs naturally with it; the interquartile range, the width of the middle 50%, pairs naturally with the median and shrugs off outliers the same way. The range is a fine sanity check and a poor conclusion.

Some quantities cannot be averaged by adding and dividing, no matter how clean the data is. Growth rates are the classic case. An investment gains 100% in year one and loses 50% in year two. The arithmetic mean of +100% and −50% is +25% a year — yet the value is exactly where it started. The honest answer comes from the geometric mean: multiply the growth factors and take the nth root, √(2.00 × 0.50) = 1, an average of 0%. Whenever the values multiply rather than add, the geometric mean is the correct average. Rates over a fixed distance need a third kind. Drive somewhere at 60 mph and return along the same road at 30 mph, and the average speed for the trip is not 45 mph. You spend twice as long on the slow leg, so it counts twice as heavily: the real figure is 40 mph — the harmonic mean. And averaging averages is its own trap. The mean of a class of 30 and a class of 10 is not the plain average of their two means unless you weight each by its size. Adding percentages, rates or averages that rest on different bases produces a number that looks reasonable and is simply wrong.