The standard error (SE) tells you how far a sample statistic, such as the mean, is likely to fall from the true value for the whole population. For the mean, you calculate it with the formula SE = s / √n, where s is the sample standard deviation and n is the sample size. A small standard error signals that your sample mean is a precise estimate of the population mean; a large one warns that it could be well off the mark.
Standard error is a cornerstone of inferential statistics. It underpins confidence intervals, hypothesis tests and regression output, and it appears everywhere from psychology and medicine to economics and econometrics. This guide explains what the standard error is, how it differs from the standard deviation, why it shrinks as your sample grows, and how to calculate it step by step — with two fully worked examples.
Understanding the Standard Error in Research
The standard error is the standard deviation of a sampling distribution — the distribution you would get if you drew many separate samples from the same population and recorded a statistic (such as the mean) from each one. The most common version is the standard error of the mean (SEM), which describes how much sample means typically vary around the true population mean.
Intuitively, the smaller the standard error, the more closely your sample mean is expected to track the population mean, and the more representative your sample is likely to be:
“The standard error of the mean (SEM) measures how far the sample mean (average) of the data is likely to be from the true population mean. The SEM is always smaller than the standard deviation.” — Investopedia, “Standard Error of the Mean (SEM)”
The Standard Error of the Mean Formula
When the population standard deviation (σ) is known, the standard error of the mean is:
SE = σ / √n
Where:
- σ — the population standard deviation
- n — the sample size (the number of observations in the sample)
In practice the population standard deviation is rarely known, so researchers estimate it using the sample standard deviation (s). The formula then becomes:
SE = s / √n
Where:
- s — the sample standard deviation
- n — the sample size
The same logic extends to other statistics. Other commonly reported standard errors include:
- the standard error of a proportion;
- the standard error of a regression coefficient;
- the standard error of the difference between two means;
- the standard error of the median.
Before going further, it is worth clearing up the single most common point of confusion: the difference between the standard error and the standard deviation.
Get statistical analysis help at an affordable price
We have:
- An expert statistician will complete your work
- Rigorous quality checks
- Confidentiality and reliability
- Any statistical software of your choice
- Free Plagiarism Report
Standard Error vs. Standard Deviation
Although they look similar and are closely related, the standard deviation and the standard error answer different questions. The standard deviation describes how spread out the individual data points are around the mean of a single dataset. The standard error describes how much a sample statistic (such as the mean) would vary from sample to sample — in other words, how precisely that statistic estimates the population value.
The two are linked directly through the sample size, because the standard error of the mean is simply the standard deviation divided by the square root of n:
SE = s / √n
Since √n is always greater than 1 for any sample larger than one observation, the standard error of the mean is always smaller than the standard deviation. Here is a side-by-side comparison:
| Feature | Standard Deviation (s) | Standard Error (SE) |
|---|---|---|
| What it measures | Spread of individual data points around the mean | Precision of the sample mean as an estimate of the population mean |
| Type of statistic | Descriptive | Inferential |
| Effect of larger sample | Tends towards the true population SD (stabilises) | Gets smaller, approaching zero |
| Formula (for the mean) | s = √[ Σ(xᵢ − x̄)² / (n − 1) ] | SE = s / √n |
| Relative size | Larger | Always smaller than s (for n > 1) |
A useful rule of thumb: report the standard deviation when you want to describe the variability within your data, and the standard error when you want to convey how confident you are in an estimate such as the mean. For more on summarising spread, see our guide to the measures of variability and the related concept of variance.
Why is Standard Error Important?
When a researcher draws a sample from a population, the sample mean is used to estimate the unknown population mean. Because a sample is only a fraction of the population, the sample mean will almost never equal the population mean exactly — the gap between them is known as sampling error. The standard error quantifies the typical size of this gap.
Even careful probability sampling, in which every element has a known chance of selection, cannot eliminate sampling error entirely. No sample matches its population perfectly, so some random variation always remains. By calculating the standard error you can judge how trustworthy your estimate is and build the confidence intervals and hypothesis tests that formal inferential statistics rely on. It is also central to assessing statistical significance: many test statistics are calculated by dividing an estimate by its standard error.
Sample Size and the Standard Error of the Mean
The standard error and the sample size are inversely related: as one increases, the other decreases. This is why larger studies produce more precise estimates. The relationship is not linear, however — it depends on the square root of the sample size, so the gains diminish as samples grow.
The table below shows how the standard error of the mean falls as n rises, holding the standard deviation fixed at s = 10:
| Sample size (n) | √n | SE = 10 / √n |
|---|---|---|
| 4 | 2.00 | 5.00 |
| 25 | 5.00 | 2.00 |
| 100 | 10.00 | 1.00 |
| 400 | 20.00 | 0.50 |
| 1,000 | 31.62 | 0.32 |
Notice that quadrupling the sample from 25 to 100 only halves the standard error, from 2.00 to 1.00. As n grows very large, the standard error approaches zero and the sample mean closes in on the true population mean. This behaviour is guaranteed by the Central Limit Theorem, which states that, for a sufficiently large sample, the sampling distribution of the mean is approximately normal with a standard deviation equal to σ/√n — regardless of the shape of the original population.
Worked through a simple scenario: suppose a class of 60 students sits a biology test, and you draw two random samples of 30 students each. Both sample means estimate the class average, but because each uses a reasonably large slice of the class, each standard error of the mean will be modest. Had you instead drawn samples of just 5 students, the means would bounce around far more from sample to sample, and the standard error would be correspondingly larger.
Calculating the Standard Error (Worked Examples)
To calculate the standard error of the mean from raw data, follow these steps:
- Find the sample mean (x̄) — add up all the observations and divide by the number of observations, n.
- Find each deviation from the mean — subtract the mean from every observation, (xᵢ − x̄).
- Square the deviations and sum them — compute Σ(xᵢ − x̄)².
- Divide by (n − 1) to obtain the sample variance, s². Using n − 1 rather than n gives an unbiased estimate of the population variance (this is known as Bessel’s correction).
- Take the square root of the variance to get the sample standard deviation, s.
- Divide s by √n to get the standard error: SE = s / √n.
Step 1 — Mean: x̄ = (10 + 20 + 24 + 29 + 40) / 5 = 123 / 5 = 24.6
Step 2 & 3 — Squared deviations:
(10 − 24.6)² = (−14.6)² = 213.16
(20 − 24.6)² = (−4.6)² = 21.16
(24 − 24.6)² = (−0.6)² = 0.36
(29 − 24.6)² = (4.4)² = 19.36
(40 − 24.6)² = (15.4)² = 237.16
Σ(xᵢ − x̄)² = 491.2
Step 4 — Sample variance: s² = 491.2 / (5 − 1) = 491.2 / 4 = 122.8
Step 5 — Sample standard deviation: s = √122.8 ≈ 11.08
Step 6 — Standard error: SE = 11.08 / √5 = 11.08 / 2.236 ≈ 4.96
Here is a second example using a larger dataset, so you can see the same method applied again.
Step 1 — Mean: x̄ = 146 / 10 = 14.6
Steps 2–3 — Sum of squared deviations: Σ(xᵢ − x̄)² = 58.4
Step 4 — Sample variance: s² = 58.4 / (10 − 1) = 58.4 / 9 ≈ 6.49
Step 5 — Sample standard deviation: s = √6.49 ≈ 2.55
Step 6 — Standard error: SE = 2.55 / √10 = 2.55 / 3.162 ≈ 0.81
Notice how the larger, less spread-out dataset in Example 2 produces a far smaller standard error (0.81) than Example 1 (4.96) — a direct consequence of its smaller standard deviation and larger sample size.
Standard Error and Confidence Intervals
One of the most important uses of the standard error is building a confidence interval — a range of plausible values for the population mean. A confidence interval for the mean takes the form:
x̄ ± (critical value × SE)
For a 95% confidence interval based on a large sample (using the normal distribution), the critical value is approximately 1.96. For smaller samples you would use the corresponding value from the t-distribution instead. Returning to Example 2, where x̄ = 14.6 and SE ≈ 0.81, an approximate 95% confidence interval (using 1.96) would be:
14.6 ± (1.96 × 0.81) = 14.6 ± 1.59 = (13.01, 16.19)
In plain terms, we can be about 95% confident that the true population mean lies between roughly 13.0 and 16.2. Because the interval width is driven by the standard error, a smaller SE — from a larger sample or less variable data — produces a narrower, more informative interval. For a full walkthrough of how these ranges are constructed and interpreted, see our guide on explaining confidence intervals.
Struggling with standard errors and confidence intervals?
ResearchProspect to the rescue!
Our expert statisticians can run, check and interpret your analysis — explore our statistical analysis service.
