Statistics Calculator

How to Calculate Standard Deviation Step by Step

First find the mean of your dataset. Then, for each value, subtract the mean and square the result — this removes negative signs and emphasizes larger deviations. Sum all those squared differences, then divide by n (population) or n-1 (sample) to get the variance. Finally, take the square root of the variance to get the standard deviation, back in the same units as your original data.

$$\sigma = \sqrt{\frac{\sum (x_i - \bar{x})^2}{n}} \qquad s = \sqrt{\frac{\sum (x_i - \bar{x})^2}{n-1}}$$

For the dataset {3, 4, 5, 6, 7, 8}, the mean is 5.5 and the sum of squared differences from the mean is 17.5. Dividing by n = 6 gives a population variance of about 2.92 (standard deviation ≈ 1.71); dividing by n − 1 = 5 gives a sample variance of 3.5 (standard deviation ≈ 1.87) — slightly larger, as Bessel's correction always produces.

Sample vs. Population Standard Deviation

If your dataset represents an entire population — every student in a specific class, every transaction in a specific month — use population standard deviation (divide by n). If your dataset is a sample drawn from a larger population you're trying to draw conclusions about — a survey of 200 people representing a country — use sample standard deviation (divide by n-1). The n-1 correction (Bessel's correction) compensates for the tendency of a sample to underestimate the true population variance.

Mean vs. Median — When to Use Each

The mean (average) uses every value in the calculation, which makes it sensitive to outliers — a single very large or very small value can pull it noticeably. The median, the middle value when sorted, is unaffected by how extreme the outliers are, only by how many values sit above versus below it. That's why income and home price statistics are usually reported as medians: a handful of very high earners or expensive homes would otherwise distort a mean.

For example, the dataset {1, 2, 3, 4, 100} has a mean of $\frac{1+2+3+4+100}{5} = 22$, pulled far above where most of the data actually sits, because of the single outlier. The median is just 3 — the middle value when sorted — which better represents where a "typical" value in this dataset actually falls.

How to Interpret Standard Deviation

A small standard deviation relative to the mean means your data points cluster tightly together — consistent, predictable. A large standard deviation means more spread and variability. In a roughly normal (bell-curve) distribution, about 68% of values fall within one standard deviation of the mean, and about 95% fall within two — a useful rule of thumb for gauging how typical or unusual a given value is.

A Brief History of Standard Deviation

The mean and simpler measures of spread date back centuries, but the term "standard deviation" and its modern symbol were introduced by the influential statistician Karl Pearson in a paper published in 1894, as part of his broader work formalizing statistics as a rigorous mathematical discipline. Pearson built on earlier work by figures like Carl Friedrich Gauss, whose study of measurement errors in astronomy in the early 1800s established much of the mathematical foundation for the normal (bell-curve) distribution that standard deviation is most closely associated with. The concept of Bessel's correction — dividing by n-1 instead of n when working with a sample — is named for the German astronomer and mathematician Friedrich Bessel, and it addresses a subtle but important bias: a sample's spread systematically underestimates the true population's spread unless this correction is applied.

Common Statistics Mistakes

Using population standard deviation (dividing by n) when the data is actually a sample is one of the most common errors, and it produces a value that's slightly too small since it skips Bessel's correction. Reporting a mean without checking for outliers is another frequent issue, since a small number of extreme values can make the "average" a poor description of what's actually typical — the median-vs-mean example above illustrates this directly. Assuming a dataset follows a normal distribution without checking is also common; the 68-95 rule of thumb only applies to roughly bell-shaped data, and can be badly misleading for skewed or multimodal datasets.

Statistics Terms You Should Know

Variance — the average of the squared differences from the mean; standard deviation is simply its square root, expressed in the original units of the data.

Outlier — a data point that differs substantially from the rest of the dataset, capable of disproportionately influencing the mean and standard deviation.

Mode — the most frequently occurring value in a dataset; unlike mean and median, a dataset can have more than one mode, or none at all.

Normal Distribution — a symmetric, bell-shaped distribution where most values cluster near the mean, and the 68-95-99.7 rule describes how data spreads across standard deviations.

Range — the difference between the largest and smallest values in a dataset, the simplest (and most outlier-sensitive) measure of spread.