Standard Error Calculator
Calculate the standard error of the mean (SEM = s/√n) or the standard error of estimate for regression (SEE = √(SSres/(n−k−1))) — with complete step-by-step working shown for every calculation.
Standard Error of the Mean
Standard Error of Estimate
Measures how much the sample mean is expected to differ from the true population mean. Divide the sample standard deviation by the square root of the sample size.
Example: s = 10, n = 25 → SE = 10/√25 = 10/5 = 2
Used for categorical/binomial data — measures uncertainty of a sample proportion p as an estimate of the population proportion.
Example: p = 0.4, n = 200 → SE = √(0.4×0.6/200) = √0.0012 ≈ 0.0346
Combines the standard errors of two independent samples — the core denominator of a two-sample t-test.
Example: s₁=5, n₁=30, s₂=6, n₂=25 → SE = √(25/30 + 36/25) = √(0.833+1.44) ≈ 1.508
Measures the average vertical distance between observed data points and a fitted regression line. Smaller SEE means the regression line fits the data more closely.
Example: SSres = 120, n = 20, k = 1 → SEE = √(120/18) ≈ 2.582
Standard error is multiplied by a critical value (z* or t*) to build a margin of error and, from that, a confidence interval around a sample estimate.
Explore this relationship fully with our Margin of Error Calculator.
Standard error is the denominator's core component in every t-statistic formula. A smaller SE produces a larger, more significant t-value for the same difference.
See it applied directly in our T-Test Calculator.
| Property | Standard Deviation (SD) | Standard Error (SE) |
|---|---|---|
| What it measures | Spread of individual data points | Precision of a sample statistic (e.g. mean) |
| Formula | √(Σ(x−x̄)²/(n−1)) | SD / √n |
| Changes with n? | Roughly stable | Shrinks as n increases |
| Used for | Describing variability in data | Confidence intervals, t-tests, margin of error |
| Units | Same as raw data | Same as raw data, but smaller |
| Statistic | Formula | Notes |
|---|---|---|
| SEM | s/√n | Click to load example |
| SE of proportion | √(p(1−p)/n) | Binomial data |
| SE of difference of means | √(s₁²/n₁+s₂²/n₂) | Two-sample t-test |
| SEE (regression) | √(SSres/(n−k−1)) | Click to load example |
| Margin of error | critical value × SE | Confidence intervals |
What Is Standard Error
This standard error calculator computes the standard error of the mean (SEM) and the standard error of estimate for regression, showing complete step-by-step working for every result. Standard error (SE) is a measure of how much a sample statistic — most commonly the sample mean — is expected to vary from the true population parameter purely due to random sampling variability.
In plain terms: if you took many different samples from the same population and calculated the mean of each one, those sample means would not all be identical — they would scatter around the true population mean. Standard error quantifies exactly how much that scatter is expected to be.
A small standard error means your sample mean is likely very close to the true population mean — your estimate is precise. A large standard error means there is more uncertainty in how well your sample represents the population.
Why Standard Error Matters
- It is the building block of confidence intervals — how wide or narrow your estimate's range is
- It is the denominator of every t-statistic and z-statistic — the foundation of hypothesis testing
- It determines the margin of error reported in polls, surveys, and experiments
- In regression, the standard error of estimate tells you how well your model's predictions match observed data
Standard Error vs Standard Deviation — The Difference
This is the single most-searched conceptual confusion in introductory statistics, and for good reason — the two terms sound alike, use similar formulas, and are often used loosely in casual conversation. But they answer fundamentally different questions.
Standard Deviation (SD) — Spread Within a Sample
Standard deviation describes how spread out individual data points are around the mean, within one dataset. If you measured the heights of 30 students, the standard deviation tells you how much any individual student's height typically differs from the average height in that group.
Standard Error (SE) — Precision of an Estimate
Standard error describes how much the sample mean itself would vary if you repeated the sampling process many times. It is not about the spread of individual data points — it is about the reliability of the summary statistic (the mean) as an estimate of the true population mean.
| Aspect | Standard Deviation | Standard Error |
|---|---|---|
| Answers the question | "How spread out is my data?" | "How precise is my sample mean?" |
| Depends on sample size? | Barely changes as n grows | Shrinks as n grows (÷√n) |
| Used in | Descriptive statistics, z-scores | Confidence intervals, t-tests, margin of error |
| Formula relationship | SD is the numerator | SE = SD ÷ √n |
Key insight: Standard deviation does not shrink meaningfully as you collect more data — it reflects the natural variability of the population. Standard error, on the other hand, always shrinks as sample size increases, because more data gives you a more reliable estimate of the true mean. This is why researchers chase larger samples: not to reduce variability in the data itself, but to reduce the uncertainty in their estimate of the mean.
Example — Same Data, Two Different Numbers
- Sample: 10 test scores with standard deviation s = 12
- Standard deviation = 12 (describes how spread out individual scores are)
- Standard error = 12/√10 ≈ 3.79 (describes how precise the sample mean is as an estimate of the true average score)
- Notice: SE is always smaller than SD (for n > 1) — this is expected and correct.
How to Calculate Standard Error — Step-by-Step
Follow this process to calculate the standard error of the mean reliably from raw data or summary statistics:
- Step 1 — Gather your sample data or identify the sample standard deviation (s) and sample size (n) if already known.
- Step 2 — Calculate the sample mean (if starting from raw data): x̄ = Σx / n
- Step 3 — Calculate the sample standard deviation: s = √( Σ(xᵢ − x̄)² / (n−1) ) — note the n−1 denominator (Bessel's correction) for sample data.
- Step 4 — Take the square root of the sample size: √n
- Step 5 — Divide: SE = s / √n
Worked Example — From Raw Data
- Data: 12, 15, 14, 18, 20, 17, 13, 16 (n = 8)
- Mean: x̄ = (12+15+14+18+20+17+13+16)/8 = 125/8 = 15.625
- Sum of squared deviations: Σ(xᵢ−x̄)² ≈ 47.875
- Sample variance: 47.875/(8−1) ≈ 6.839
- Sample standard deviation: s = √6.839 ≈ 2.615
- Standard error: SE = 2.615/√8 ≈ 2.615/2.828 ≈ 0.925
Standard Error of the Mean vs Standard Error of Estimate
Both quantities are called "standard error," but they measure the precision of two very different things — a sample mean versus a regression prediction.
Standard Error of the Mean (SEM)
Used when you have a single sample and want to know how precisely its mean estimates the population mean. Formula: SEM = s/√n. Appears in confidence intervals for a mean and one-sample t-tests.
Standard Error of Estimate (SEE) — Regression
Used in linear regression to measure how far, on average, the observed data points fall from the fitted regression line. A smaller SEE means the regression line predicts the data more accurately. Formula: SEE = √(SSres/(n−k−1)), where SSres is the sum of squared residuals, n is the number of observations, and k is the number of predictor variables.
| Aspect | SEM | SEE |
|---|---|---|
| Measures precision of | The sample mean | The regression line's predictions |
| Formula | s/√n | √(SSres/(n−k−1)) |
| Context | Descriptive / inferential stats | Regression analysis |
| Degrees of freedom | n−1 | n−k−1 |
Standard error is the denominator's core component in every t-statistic formula — explore this directly with our T-Test Calculator. And remember: margin of error = critical value × standard error — see this relationship in action with our Margin of Error Calculator.
Worked Examples
1. Standard Error of the Mean — Direct Formula
- Given: s = 8, n = 16
- SE = s/√n = 8/√16 = 8/4 = 2
2. Standard Error of the Mean — Effect of Sample Size
- Given: s = 15, n = 30 → SE = 15/√30 ≈ 2.739
- Now quadruple the sample: n = 120 → SE = 15/√120 ≈ 1.369
- Notice: quadrupling n only halves SE — because SE scales with 1/√n, not 1/n.
3. Standard Error of Estimate — Simple Linear Regression
- Regression fitted to 20 data points, one predictor (k=1)
- Sum of squared residuals: SSres = 144
- Degrees of freedom: n−k−1 = 20−1−1 = 18
- SEE = √(144/18) = √8 ≈ 2.828
4. Standard Error of Estimate — Multiple Regression
- Regression with 50 observations and 3 predictors (k=3)
- SSres = 620
- Degrees of freedom: 50−3−1 = 46
- SEE = √(620/46) ≈ 3.673
5. From Raw Paired Data to SEE
- Data pairs: (1,2.1), (2,3.9), (3,6.2), (4,7.8), (5,10.1)
- Fit y = a + bx by least squares → slope b ≈ 1.99, intercept a ≈ 0.03
- Compute residuals for each point, square and sum them → SSres
- n = 5, k = 1 → degrees of freedom = 3
- SEE = √(SSres/3) — use the calculator above to see this computed live from the raw pairs.