Skip to content
UniKit

Statistics calculator

Mean, median, mode, extremes, variance and standard deviation, quartiles and IQR, geometric and harmonic mean, skewness and kurtosis — from data separated by commas, spaces or new lines.

Runs in your browserEvery computation happens in your browser — your data never leaves this device.

4
Count8
Sum40
Mean5
Median4.5
Mode4
Minimum2
Maximum9
Range7
Population variance4
Sample variance4.5714
Population standard deviation2
Sample standard deviation2.1381
First quartile (Q1)4
Third quartile (Q3)6
Interquartile range (IQR)2
Geometric mean4.6032
Harmonic mean4.2018
Skewness0.6563
Kurtosis (excess)-0.2188

What this tool does

  • After an experiment or an A/B test, get mean, median, standard deviation and quartiles in one pass instead of building a wall of spreadsheet formulas.
  • When outliers make the choice between mean and median unclear, both are shown side by side along with the mode and a frequency table.
  • Check hand-computed skewness, kurtosis, geometric mean and harmonic mean for homework or teaching material.
  • Audit the convention you are using: population variance (divide by n) and sample variance (divide by n-1) are listed next to each other so the denominator is never a guess.

Example

Input

2, 4, 4, 4, 5, 5, 7, 9

Output

Count 8, sum 40, mean 5, median 4.5, mode 4, min 2, max 9, range 7
Population variance 4, sample variance 4.5714, population SD 2, sample SD 2.1381
Q1 = 4, Q3 = 6, IQR = 2, geometric mean 4.6032, harmonic mean 4.2018, skewness 0.6563, kurtosis -0.2188

Four decimal places by default. Quartiles use Tukey hinges: the eight values split into two groups of four, whose medians are 4 and 6.

Frequently asked questions

How do I separate the values — can I paste a column from a spreadsheet?

Commas, Chinese commas, enumeration commas, spaces, tabs and line breaks all work as separators, so pasting a column straight out of Excel is fine. Anything unparseable such as "n/a" is reported rather than silently skipped.

Population or sample variance — which one applies to me?

Use population variance (divide by n) when the data is the entire group you care about, and sample variance (divide by n-1) when the data is a sample drawn from a larger population, since only the latter is an unbiased estimator. Both are computed so you can pick per case.

Why does the mode sometimes say "No mode"?

If every value appears exactly once there is no mode, and the tool returns an empty set. When several values tie for the highest frequency they are all listed — the mode of 1, 1, 2, 2 is 1, 2.

Why do the geometric and harmonic means show a dash?

Both are only defined when every value is positive, so a single 0 or negative value makes them unavailable. Small samples matter too: skewness needs at least 3 values, kurtosis at least 4, and the sample variance and standard deviation at least 2.

How do I control the number of decimals?

Use the decimal places slider — 0 to 15, four by default. Summation uses Kahan compensated addition, which pulls accumulation error in cases like 0.1 + 0.2 + 0.3 back to the floating point limit, so a mean of 5 prints as 5 and not 5.000000000000001.

Keywords:statisticsmeanmedianmodevariancestandard deviationquartileiqrkurtosisskewness统计统计计算器均值中位数众数方差标准差四分位距偏度峰度

Related tools