Sample vs population standard deviation#
Standard deviation measures how spread out values are around the mean. If your list is the entire population you care about (every student in a class, every day in a fixed campaign window), use the population formula: divide the sum of squared deviations by n, then take the square root. If the list is a sample used to estimate a larger unknown population, use the sample formula: divide by n − 1 (Bessel’s correction) so the estimate is less biased low. This calculator shows both so you do not have to guess which “standard deviation calculator” mode a generic form silently picked.
A common homework trap is treating a classroom dataset as a sample when the question defines that class as the whole population — or the reverse when a survey of 30 people is meant to stand in for a city. The formulas only differ in the denominator, but the interpretation differs a lot: population σ describes this exact set; sample s estimates an unseen larger set. When in doubt for statistics homework, read whether the problem says “population” or “sample”; for analytics work, ask whether you will ever see more data from the same process.
How the calculation uses the mean#
Every path starts the same way: compute the arithmetic mean x̄ = Σxᵢ / n. Then for each value compute (xᵢ − x̄)², sum those squared deviations, and divide — by n for population variance σ², or by n − 1 for sample variance s². Standard deviation is the square root of the matching variance. A “standard deviation calculator using mean” is not a different method; the mean is the center the formula always measures distance from. Optional steps in the tool spell that sequence out with your numbers plugged in.
Squaring before averaging is what makes large outliers dominate the result: a point twice as far from the mean contributes four times as much to the sum of squares. That is intentional — standard deviation is sensitive to tails — and it is why a single typo (an extra zero) can inflate σ dramatically. If your pasted list looks right but the stdev looks absurd, scan min/max on this page first; they are the fastest sanity check before trusting the spread.
Matching Excel STDEV.S and STDEV.P#
In Excel and Google Sheets, STDEV.S / STDEV match this tool’s sample standard deviation, and STDEV.P / STDEVP match the population value. Paste the same column of cells here (commas or newlines are fine) and you should land on the same figures, aside from display rounding. If your spreadsheet result disagrees, check whether the sheet used the sample or population function, and whether blank cells or text slipped into the range.
Copying from a spreadsheet often introduces trailing blanks or locale decimal commas. This parser accepts commas, spaces, newlines, and semicolons as separators and requires each token to be a finite number — so 1; 2; 3 and a newline-separated Excel paste both work, while a cell containing N/A will surface a clear parse error instead of quietly dropping the row the way some sheet ranges do.
Standard deviation and the normal distribution#
For data that follow a roughly normal (bell-curve) distribution, about 68% of values fall within one standard deviation of the mean, about 95% within two, and about 99.7% within three — the empirical 68–95–99.7 rule. This page calculates the spread itself; it does not fit a curve or return z-table probabilities. Use the stdev you get here as the σ (or s) you would plug into a normal-distribution or z-score workflow when the normality assumption is reasonable for your data.
Not every dataset is normal. Skewed revenue, bounded percentages, or multimodal survey scores can have a well-defined standard deviation without the 68–95–99.7 percentages applying cleanly. Treat those percentages as a rule of thumb for mound-shaped, symmetric data — and treat this calculator as the place to get σ or s first, before any distributional claim.