Standard Error Calculator
The precision of an ESTIMATE: SE of the mean from s and n, SE of a proportion, and SE of a difference — with the card that keeps SD and SE from being confused, because one describes people and the other describes your answer.
Standard Error Calculator
Results recalculate instantly on every keystroke. Nothing you type is transmitted.
What this result does not account for
- Simple random samples — no clustered or weighted designs
- Normal-reading of SE leans on the central limit theorem at small n
In short: A sample of n = 36 with s = 12: the standard error of the mean is 12/√36 = 2 — the estimate’s own wobble across repeated samples, not the spread of the 36 people (that is the 12). A proportion p̂ = 0.5 at n = 400 carries SE = √(0.25/400) = 0.025 — the number every margin of error multiplies. And the difference of two means (s₁ = 10, n₁ = 50; s₂ = 15, n₂ = 60) has SE = √(100/50 + 225/60) = 2.397916. The doctrine card: SD answers “how spread out are the data?”, SE answers “how sure are you about the average?” — quoting one when you mean the other is the oldest error in applied statistics.
Formula
SE = s/√n · SE(p̂) = √(p̂(1−p̂)/n) · SE(diff) = √(s₁²/n₁ + s₂²/n₂)
Three flavours, one idea: an estimate wobbles by the data’s spread divided by the square root of how much of it you collected. The margin-of-error page multiplies these by z*; the t page divides by them.
Worked Example
- Give the sample size and the sample SD for the mean’s SE.
- Give a proportion for the second card; a second sample’s size and SD for the difference.
- Watch the √n card re-price precision at 4n and n/4 — the economics of every study.
- Quote SE only when you mean the estimate’s wobble — the doctrine card holds the line.
Defaults: SE mean 12/√36 = 2; SE proportion √(0.25/400) = 0.025; SE difference 2.397916. Quadruple n and the SE halves — priced live on your own numbers.
Strengths & Limits Of This Model
Where this engine is strong
- All three estimate flavours priced on one panel
- The √n law re-priced live at 4n and n/4 on your numbers
Where it stops
- No bootstrap or robust SE variants
- No clustered/stratified SEs — design effects out of scope
Practical Use Cases
Methods sections
report the estimate’s precision
Survey QC
the proportion’s wobble before any margin
Teaching
SD vs SE, settled with your own numbers
Methodology & Editorial Standards
Computation runs in IEEE-754 double precision at full internal precision; rounding to two decimal places occurs strictly at the display layer, so no cumulative drift enters the result. All monetary outputs use accounting presentation — grouped thousands, two decimals, negatives in parentheses — so figures can be transcribed directly into a model or working paper. Division-by-zero and out-of-domain inputs return an em-dash rather than a misleading number.
This engine was reconciled against an independent reference implementation and hand-verified for the worked example above before release. Our full five-stage review process is published on the About Us page.
Disclaimer. This calculator is provided for informational and modelling purposes only and does not constitute financial, tax, legal, medical, or engineering advice. Verify all figures with a qualified professional before acting on them.
Standard Error Calculator — 8 Expert FAQs
8 analyst-written answers to the questions practitioners actually ask — optimised for voice and answer-engine retrieval.
What is the difference between SD and SE?
SD describes the DATA: how far people sit from the average. SE describes your ESTIMATE: how far the average of a different sample would sit from this one. Same units, different questions — and swapping them in a paper is the oldest error in applied statistics. SE = SD/√n always ties them.
Why does everything divide by √n?
Averages cancel noise: independent wobbles partially cancel when averaged, and the cancellation is exactly √n. The √n card re-prices it live — quadruple the sample and the SE halves; quarter it and the SE doubles. Precision is bought by the square, the same law the margin page teaches.
Why does the proportion use p̂(1−p̂)?
Because a proportion’s variance comes from the Bernoulli coin underneath: p(1−p). It peaks at p = 0.5 — coin-flip splits wobble most — and shrinks toward the extremes, which is the same geometry the margin-of-error page plans on when it defaults to 0.5.
When do I need the SE of a difference?
Whenever two groups are compared: treatment minus control, variant A minus variant B. Each mean wobbles on its own, and independent wobbles ADD in variance — hence the plus under the square root, 2.397916 on the defaults. The t page divides the observed gap by exactly this.
Should I use s or σ here?
Whatever you actually have. From a SAMPLE, s (the n − 1 divisor) is the honest input and the SE inherits its small-sample noise; a KNOWN σ belongs to the known-σ flavour the margin page prices. What you may not do is compute s and call it σ — the divisor split the variance page owns is the difference.
Can SE be zero?
Only when the data have no spread (s = 0) — and then the SE card reports a precise answer built on a degenerate sample: every observation identical, so the estimate cannot wobble but nothing was learned about anyone else. The page prints it with that warning attached.
How big should n be before SE means much?
SE’s FORMULA runs at any n, but its normal-reading (the ±2 SE rule of thumb) leans on the central limit theorem, which needs n in the double digits for skewed data. The t machinery on the hypothesis page exists precisely for small-n honesty.
Why three flavours on one page?
Because they are one idea wearing three hats: spread over root-count. Splitting them across pages would hide that the mean, proportion and difference estimates all wobble by the same law — and the whole point of this page is the law, with your numbers in all three seats.