Mean, Median, Mode & Range Calculator

Enter a list of numbers in the input box. Separate observations with commas, spaces, semicolons, tabs, or line breaks.

The calculator immediately returns the mean, median, mode, and range. It also shows the count, sum, minimum, and maximum.

Use the decimal-place selector to control visible rounding. The underlying observations and calculated results are not changed.

Use Copy to copy the summary or Frequency CSV to download the complete exact-value frequency table.

Mean, median, and mode are measures of location or central tendency, but they do not define the centre in the same way.

Mean uses every observation. Median uses the sorted middle position. Mode uses repetition.

Range is a measure of spread rather than centre. It reports the distance between the smallest and largest observations.

Looking at all four together gives more context than selecting one number without considering the shape of the data.

This dataset shows why a complete summary is more useful than one average.

The four smaller values sit between 2 and 5, but the value 100 raises the mean to 22.8.

The median remains 4 because it depends on sorted position rather than the distance to the largest observation.

Every value appears once, so the calculator reports no mode. The range is 98.

The arithmetic mean is the sum of all observations divided by the number of observations.

For 4, 7, and 10, the sum is 21 and the count is 3, so the mean is 7.

Every value contributes to the result. Increasing one value by 30 increases the total by 30 and therefore changes the mean by 30 divided by the count.

The calculator uses compensated summation to reduce avoidable floating-point error when adding long datasets or values with very different magnitudes.

The median is found after the observations are sorted from lowest to highest.

For an odd number of observations, the median is the single middle value. In 2, 4, 9, the median is 4.

For an even number of observations, the median is the arithmetic average of the two middle values. In 1, 4, 7, 10, the median is 5.5.

The median is often less affected than the mean by a small number of extremely high or low values.

The mode is the exact value or values with the highest repeated frequency.

In 1, 2, 2, 3, the mode is 2 because it occurs twice while every other value occurs once.

In 1, 1, 2, 2, 3, the modes are 1 and 2 because both occur twice. This is a tied or multimodal result.

When every distinct value occurs once, the calculator reports No mode.

Frequency counting happens after each token has been converted into a number.

Inputs such as 1, 1.0, and 1e0 all become the numerical value 1 and therefore share one frequency count.

Nearby values remain different. For example, 1.2 and 1.2001 are separate observations even when the selected display precision makes both appear as 1.2.

The calculator does not round values into groups before finding the mode.

Range equals the maximum observation minus the minimum observation.

For 3, 5, 8, and 11, the range is 11 − 3, or 8.

Range remains in the same unit as the original observations.

It is easy to calculate but uses only two values. It does not describe how the remaining observations are distributed between the endpoints.

Mean and median can be similar in balanced data and very different in skewed data.

In the dataset 2, 3, 4, 5, 100, the mean is 22.8 while the median is 4.

Neither result is mathematically wrong. The mean describes the arithmetic balance point, while the median describes the middle sorted position.

For salaries, property prices, response times, and other long-tailed data, reporting both values can prevent one measure from hiding an important feature of the distribution.

The same statistic can tell a different story depending on the distribution of observations.

A repeated value can become the mode without being close to the mean. Two datasets can also share the same range while having completely different centres and frequency patterns.

Use the table to compare how the four measures respond to symmetry, outliers, repetition, tied modes, and negative values.

Count is the number of observations included in the calculation.

Sum is the total of all observations and is used directly in the arithmetic mean.

Minimum and maximum are the smallest and largest entered values after numerical parsing.

These supporting results make the main statistics easier to verify and help reveal incomplete or incorrectly entered data.

A frequency table records how often each exact numerical value occurs.

Relative frequency expresses that count as a percentage of the full dataset.

For example, if the value 4 appears 3 times in a dataset of 12 observations, its frequency is 3 and its percentage is 25%.

The CSV export includes every distinct value, its count, and its percentage, even when the interface displays only the main summary statistics.

These measures are useful for quickly summarizing test scores, prices, salaries, measurements, response times, delivery times, survey answers, sales figures, sports results, and experimental observations.

The most useful measure depends on the question, the data type, and the shape of the distribution.

A calculation can be numerically correct while still being misleading when observations use inconsistent units, definitions, periods, or collection methods.

Confirm that every observation represents the same quantity and unit. Do not combine metres with centimetres, monthly values with annual values, or scores from different scales without converting them first.

Check for copied headings, blank cells, placeholders, duplicated records, missing periods, impossible measurements, and values entered with the wrong decimal separator.

Decide how missing values should be handled before using the calculator. A missing observation is not automatically zero.

Investigate apparent outliers before deleting them. An extreme value may be a data-entry error, but it may also be a valid and important observation.

The decimal-place setting changes only the text shown in the result.

Calculations and frequency counts use the parsed numerical values rather than the visibly rounded versions.

Two distinct values can therefore look identical at a low display precision while remaining separate in the frequency CSV.

Choose enough decimal places to match the precision and purpose of the original measurements rather than displaying digits the source data does not support.

Scientific notation is useful for very large or very small observations. For example, 1e3 is parsed as 1,000 and 2.5e-4 is parsed as 0.00025.

Negative values are supported. Their numerical ordering, sum, mean, median, frequencies, and range are calculated normally.

A negative range is impossible because the minimum is always subtracted from the maximum.

A comma is treated as a boundary between observations.

Entering 1,234 therefore creates the two values 1 and 234 rather than the single value one thousand two hundred thirty-four.

Enter 1234 without a grouping comma when it is one observation.

This rule keeps pasted lists predictable and avoids guessing whether a comma separates values or groups digits.

Calculations use JavaScript double-precision floating-point numbers.

Compensated and scaled summation reduce common accumulation and intermediate-overflow problems, but they do not provide exact arbitrary-precision decimal arithmetic.

Values and results outside JavaScript's finite numerical range are rejected rather than displayed as Infinity or NaN.

The 50,000-value limit bounds local parsing, sorting, frequency counting, memory use, and CSV generation so one unusually large paste does not monopolize the page.

These measures provide a compact descriptive summary but do not show the complete distribution.

Use variance or standard deviation when spread across all observations matters.

Use quartiles, percentiles, or the interquartile range when position and robust spread are important.

Use charts when shape, clusters, gaps, skewness, or several peaks need to be seen.

Use inferential statistical methods when the goal is to draw conclusions about a wider population rather than only describe the entered values.

Parsing, sorting, frequency counting, statistical calculation, copying, and CSV generation run entirely in the browser.

The entered dataset is not uploaded, added to a public URL, connected to an account, or permanently stored by the tool.

Clipboard access is requested only when Copy is used. The CSV file is generated locally when requested.