In Excel: use =SKEW(range) to measure asymmetry — positive means a long right tail — and =KURT(range) to measure how heavy the tails are.
On this page8
Syntax
Arguments
| Argument | required / optional | Description |
|---|---|---|
number1 | required | The first value or range. At least three values are needed for SKEW, four for KURT. |
number2, ... | optional | Further values or ranges. |
Related functions
Run =SKEW( over the data range before choosing a summary statistic.
Read the sign and size: beyond about ±1 the distribution is materially lopsided.
Cross-check by comparing AVERAGE against MEDIAN on the same range.
If skew is high, lead the report with the median and mention the tail explicitly.
What this does
SKEW measures how lopsided a distribution is. Zero means symmetric, positive means a long right tail (a few very large values, typical of income and response times), negative means a long left tail. KURT measures tail weight relative to a normal distribution, where positive means more extreme outliers than normal. Their practical value is as a check before applying any method that assumes normality: a skew beyond roughly ±1 says the mean is a poor summary and the median should lead instead. SKEW needs at least three values, KURT at least four. Most people learn this as a sequence of clicks and forget it by next week; learning it as a pattern instead is what lets you apply it to the next, slightly different version of the problem without starting from scratch. That is the difference this page is trying to make. The difference between a quick fix and a sheet you can trust is the extra minute you spend validating “find skewness in excel”. Start on a copy or a tiny sample, keep the affected cells visible, and compare the result with the outcome you expected before you touch the real workbook. When a formula is involved, keep the inputs labelled beside it, reference cells instead of typing values, and apply number formatting only after the result checks out. The point is a data step that keeps your analysis trustworthy, but the practical win is that someone else can open the file and understand what happened without asking you.
A worked example
Testing whether a mean is representative: =SKEW(B2:B500) returning 1.8 says the data has a long right tail and the median is the honest headline. Confirming with the two summaries: a large gap between =AVERAGE(B2:B500) and =MEDIAN(B2:B500) tells the same story. =KURT(B2:B500) above 3 warns of more extreme outliers than a normal distribution would produce. SKEW answers "is the mean lying to me", which is the question that should precede every reported average. Check the referenced cells, argument order and result type in your workbook. Start with the worked example before changing an input, then compare the changed result with the calculation you expected. An error value or a result of a different type is a reason to inspect the inputs and arguments.
In Google Sheets
Google Sheets support for this task has not been verified. Check the function, arguments and workflow in the target product before transferring the example. Matching formula notation does not establish identical behaviour. Keep this page bookmarked for the next time the same question comes up. Better still, repeat the steps once in your own workbook — doing it yourself is what turns a copied answer into something you remember. For “find skewness in excel”, keep the input cells and the result visible while you check the formula in Excel. Follow the worked example and change one input at a time. The steps explain the workbook calculation; this page does not run a live calculator for that task.
Common mistakes
- Reporting a mean on data with a skew beyond ±1, which describes almost nobody in the dataset.
- Running them on fewer than three or four values, which returns
#DIV/0!. - Treating a skew near zero as proof of normality; it rules out one departure, not all of them.
Frequently asked questions
What does a positive skew mean?
A long right tail — a few unusually large values pulling the mean above the median. Income and response times behave this way.
When should I use the median instead of the mean?
Whenever the skew is materially non-zero, roughly beyond ±1. The mean stops describing a typical case.
What does KURT tell me?
How heavy the tails are compared with a normal distribution. Positive means more extreme outliers than normal.