Average Calculator (Mean, Median, Mode)

How to use it

1. Enter your numbers. Paste or type them into the box. Separate them with commas (12, 7, 22), spaces (12 7 22), new lines, or any mix of the three.
2. Read the eight stat cards. They refresh as you type, with no button to press: count, sum, mean, median, mode, min, max and range.
3. Check the chips underneath. Every value you entered shows up as a chip, and possible outliers are highlighted in amber with a plain-English summary below them.
4. Copy or reset. Copy results puts all eight figures plus the outlier line on your clipboard. Load sample fills the box with 12, 7, 22, 7, 19, 31, 7, 14. Clear empties it.

Decimals (3.14), negatives (-5) and scientific notation (1.2e3) all work. A leading currency symbol ($, £, € or ¥) is stripped before the value is read. Anything else is skipped, and the line under the box counts how many tokens were ignored so you can spot a typo.

Watch out for thousands separators

The parser treats every comma as a separator. It always has. So 1,200 is not read as one thousand two hundred: it splits into 1 and 200, which turns one entry into two and pulls your mean down without saying a word. Strip the thousands separators before you paste (1200), or separate your values with spaces and new lines instead. The ignored-token counter will not rescue you here, because 1 and 200 are both perfectly valid numbers.

Mean, median and mode

These three are the classic measures of central tendency. They are three different answers to one question: what is a typical value?

The mean adds every number and divides by how many there are. This is the everyday “average”. Themedian sorts the numbers and takes the middle one, or the average of the two middle ones when the count is even. The mode is whichever value shows up most often. A list can have one mode, several tied modes, or none at all when every value is unique.

Why bother with all three? Because one large value can drag the mean a long way. Picture ten people in a room: nine earn £40,000 and one earns £1,000,000. The mean salary reads £136,000, a figure nobody in the room actually earns, while the median holds steady at £40,000. Reading the mean and the median side by side is the quickest skew test there is. When they sit close together your data is roughly symmetric, and when the mean sits well above the median you have a long tail pulling to the right. That gap is exactly why house prices, salaries and waiting times are almost always reported as medians rather than means, and why a mean on its own can be technically correct and still misleading.

One limitation worth naming: every value here counts equally. If you need a weighted average, where some entries carry more importance than others, this is the wrong tool. Repeat a value to give it more weight, or use a dedicated weighted average calculator.

How the outlier check works

Once you have entered at least four numbers, the calculator applies Tukey’s 1.5x IQR rule and highlights anything sitting unusually far from the rest of your data.

The mechanics are simple. Sort the list, find the first and third quartiles (Q1 and Q3), then subtract one from the other to get the interquartile range. Draw a fence at Q1 minus 1.5 × IQR and another at Q3 plus 1.5 × IQR. Anything beyond a fence gets flagged. Four values is the floor because a shorter list cannot be split into two halves that each have a median.

A flag is not a verdict. It reports that a number sits far from its neighbours in this particular list, nothing more. On perfectly normal data the fences land about 2.698 standard deviations either side of the mean, so roughly0.70 per cent of readings trip the rule by pure chance. Paste 1,000 clean measurements and you should expect about seven amber chips with nothing whatsoever wrong. Deleting flagged values because a tool coloured them yellow is how good data gets quietly thrown away.

Why your Q1 may not match Excel’s

Here is the part most average calculators leave out. There is no single agreed way to compute a quartile from a sample. Hyndman and Fan catalogued nine different definitions in The American Statistician in 1996, three based on rounding and six on interpolation, and every one of them is defensible (Hyndman & Fan, 1996).

Software simply picks a side. R, NumPy, SciPy and Julia all default to type 7. SAS defaults to type 2. This calculator uses the median-of-halves method taught in most introductory courses: split the sorted list at its midpoint, drop the middle value when the count is odd, then take the median of each half.

Those choices give different numbers, and the outlier flag can flip because of it. Take the list 13, 13, 15, 17, 81:

Same five numbers, opposite verdicts. So if a coursework brief, a lab protocol or a house style specifies a quartile method, check which one it wants before you copy a fence value across. Disagreements like this shrink fast as your sample grows, and on a few hundred values the methods usually land within a rounding error of each other.

What the other figures mean

• Count: how many valid numbers were found in the box.
• Sum: every value added together.
• Min and Max: the smallest and largest values in the list.
• Range: max minus min, a quick gauge of how spread out the data is. It only ever uses two of your numbers, so a single extreme value sets it on its own.

Frequently asked questions

Figures are rounded to six decimal places for display. For grading, lab work or anything auditable, check the critical numbers against the method your course or employer expects.