Standard error vs standard deviation explained

Q: Is standard error always smaller than standard deviation?

For the sample mean with n above 1, yes. Since sqrt(n) is greater than 1, dividing s by it makes SE smaller than s. The gap between them grows as your sample gets larger.

Q: Does the standard error get smaller with more data?

Yes. Because SE equals s/sqrt(n), a larger n shrinks the standard error, which is why bigger samples give more precise estimates. Quadrupling the sample size cuts the standard error in half.

By Jude Wallis · Published July 28, 2026

Standard deviation measures how spread out the individual values in a dataset are. Standard error measures how much a statistic like the sample mean varies from sample to sample. For the mean, SE equals s divided by the square root of n, so it is smaller than the SD.

AP Statistics: Unit 4 (topics 4.1 Sampling Distributions for Sample Means, 4.2 Constructing a Confidence Interval for a Population Mean or Population Mean Difference). In the Fall 2026 course, Topic 4.1 introduces the standard deviation of the sampling distribution of the sample mean as the population standard deviation divided by the square root of n. Topic 4.2 then defines the standard error of the sample mean as s divided by the square root of n, the form that every Unit 4 confidence interval and test for a mean uses.

The core difference between standard error and standard deviation

Standard deviation and standard error sound alike and share similar formulas, but they answer different questions.

Standard deviation (SD) measures how spread out the individual values in a single dataset are. Standard error (SE) measures how much a summary statistic, such as the sample mean, would change if you drew a new random sample.

A quick way to keep them straight: standard deviation is about the data points, and standard error is about a statistic computed from those data points.

You will see a few symbols on this page. The sample mean is written $\bar{x}$ (x-bar), the sample standard deviation is written $s$ , the population standard deviation is the Greek letter $\sigma$ (sigma), the population mean is $\mu$ (mu), and $n$ is the sample size.

Standard error vs standard deviation side by side

This table shows where the two measures agree and where they split apart.

Question	Standard deviation	Standard error of the mean
What is spread out?	The individual data values	The sample mean $\bar{x}$ from sample to sample
Symbol	$s$ (sample) or $\sigma$ (population)	SE, equal to $s/\sqrt{n}$
Formula from data	$s = \sqrt{\frac{\sum (x_i - \bar{x})^2}{n-1}}$	$\text{SE} = \frac{s}{\sqrt{n}}$
Larger sample size $n$	Stays about the same	Shrinks toward 0
Main job	Describe one distribution	Build confidence intervals and tests

Notice that the standard error is just the standard deviation divided by $\sqrt{n}$ . That single division is the whole relationship for the mean.

What standard deviation measures

Standard deviation answers the question, how far do the individual values sit from their mean, on average? A small SD means the values cluster tightly around $\bar{x}$ . A large SD means they spread out widely.

Standard deviation stays with the raw data. If your quiz scores range from 6 to 10, the SD is a fixed property of those five numbers. Collecting a second class of scores does not change the spread of this one.

Standard deviation carries the same units as your data. If you measure heights in centimeters, the SD is in centimeters too, which makes it easy to describe a typical distance from the mean. It does not shrink just because you gathered more observations, since more data does not make a spread-out variable less spread out.

The sample version divides by $n-1$ , which the AP formula sheet gives as $s = \sqrt{\frac{\sum (x_i - \bar{x})^2}{n-1}}$ . For a step by step walkthrough of that calculation, see standard deviation by hand or run the numbers with the standard deviation calculator.

What standard error measures

Standard error answers a different question, how much would the sample mean move if you took a fresh sample of the same size? Every random sample gives a slightly different $\bar{x}$ . The standard error is the standard deviation of all those possible sample means.

That collection of possible sample means has its own distribution, called the sampling distribution of $\bar{x}$ . Its center is the true mean $\mu$ , and its spread is the standard error. See sampling distributions explained for the full picture.

In practice you almost never take many samples, so you cannot watch $\bar{x}$ bounce around directly. Instead you estimate that bounce from the one sample you have, using $s/\sqrt{n}$ . The standard error is really an estimate of the standard deviation of the sampling distribution.

Because SE describes a statistic and not the raw data, it is the number you plug into confidence intervals and test statistics. It reflects the precision of your estimate, not the variability of any one measurement.

Why SE equals s over the square root of n

Averaging cancels out some of the randomness in the individual values. The more values you average, the steadier the mean becomes, so the standard error shrinks as $n$ grows.

\text{SE of } \bar{x} = \frac{s}{\sqrt{n}}

The square root matters. To cut the standard error in half, you need four times as much data, not twice as much. Going from $n = 25$ to $n = 100$ divides SE by $\sqrt{4} = 2$ .

The AP formula sheet lists the standard deviation of the sampling distribution of $\bar{x}$ as $\sigma/\sqrt{n}$ when the population SD $\sigma$ is known, and the standard error as $s/\sqrt{n}$ when you estimate the spread from the sample itself. The central limit theorem is what makes that sampling distribution roughly normal for a large enough $n$ .

Which one to report

Report the standard deviation when you are describing a single group. It tells your reader how variable the individual measurements are, such as the spread of heights in one class.

Report the standard error when you are estimating something about a larger population from your sample. Confidence intervals and hypothesis tests both use SE, which is why the AP formula sheet writes an interval as statistic $\pm$ (critical value)(standard error). Building one is covered in how to calculate a confidence interval.

If you are unsure which to use, ask what your number is about. A number about the data values is a standard deviation. A number about how precise your estimate is uses the standard error.

Common mix-ups to avoid

The two get swapped often, especially on calculator output. Here are the traps to watch for.

A calculator's 1-Var Stats screen reports $s$ (sample SD) and $\sigma$ (population SD), but not the standard error. You compute SE yourself by dividing $s$ by $\sqrt{n}$ .
Error bars on a graph can show either SD or SE. Check the caption, because SE bars look smaller and can make groups seem more separated than the data supports.
Standard error is not a mistake or a measurement error. It is the expected sample to sample variation of a statistic.

For help decoding what your calculator prints, see reading calculator output.

Computing both SD and SE from one set of quiz scores

Five students score 8, 6, 9, 7, and 10 on a 10-point quiz. Treat these as a sample. Find the sample standard deviation and the standard error of the mean.

Find the sample mean: $\bar{x} = \frac{8 + 6 + 9 + 7 + 10}{5} = \frac{40}{5} = 8$ .
Subtract the mean from each value to get the deviations: $8-8=0$ , $6-8=-2$ , $9-8=1$ , $7-8=-1$ , $10-8=2$ .
Square each deviation: $0$ , $4$ , $1$ , $1$ , $4$ . Their sum is $0+4+1+1+4 = 10$ .
Divide by $n-1 = 4$ to get the sample variance: $s^2 = \frac{10}{4} = 2.5$ .
Take the square root for the sample standard deviation: $s = \sqrt{2.5} \approx 1.5811$ .
Divide by $\sqrt{n} = \sqrt{5} \approx 2.2361$ to get the standard error: $\text{SE} = \frac{1.5811}{2.2361} \approx 0.7071$ .

The sample standard deviation is $s \approx 1.5811$ , describing the spread of the five scores. The standard error of the mean is $\text{SE} \approx 0.7071$ , describing how much the mean of five such scores would vary from sample to sample. SE is smaller than $s$ because dividing by $\sqrt{5}$ shrinks it.

Frequently asked questions

Is standard error just standard deviation divided by the square root of n?

For the mean, yes. The standard error of the sample mean equals $s/\sqrt{n}$ , where $s$ is the sample standard deviation and $n$ is the sample size. Other statistics, such as a sample proportion, have their own standard error formulas.

Is standard error always smaller than standard deviation?

For the sample mean with $n$ above 1, yes. Since $\sqrt{n}$ is greater than 1, dividing $s$ by it makes SE smaller than $s$ . The gap between them grows as your sample gets larger.

Does the standard error get smaller with more data?

Yes. Because SE equals $s/\sqrt{n}$ , a larger $n$ shrinks the standard error, which is why bigger samples give more precise estimates. Quadrupling the sample size cuts the standard error in half.

Which one do I use in a confidence interval?

The standard error. The AP formula sheet writes a confidence interval as statistic $\pm$ (critical value)(standard error), so the interval for a mean uses $s/\sqrt{n}$ , not the raw standard deviation.