How this instrument works
A confidence interval is a range built around a sample mean that is likely to contain the true population mean. '95% confidence' describes the long-run behaviour of the method: if you repeated the same sampling process many times and built a 95% interval each time, about 95% of those intervals would capture the true mean. This instrument fixes the confidence level at 95% and the normal critical value at z = 1.96, the single most widely cited two-sided critical value in applied statistics.
95% earned its status as the default not because it is mathematically special, but because it has become the field-wide convention: peer-reviewed journals, government statistics agencies, and most introductory statistics courses treat 1.96 as the standard multiplier unless a specific reason calls for something else. That convention matters in practice — a reviewer or reader who sees '95% CI' knows exactly what standard was applied without having to check.
The interval is built from three ingredients: the sample mean x-bar (your point estimate), the standard error s/√n (how much that estimate would wobble on repeat sampling), and the fixed critical value 1.96 (how many standard errors on each side capture the middle 95% of a normal distribution). Multiply the standard error by 1.96 to get the margin, then add and subtract that margin from x-bar to get the upper and lower bounds.
- Enter your sample's average into Sample mean (x-bar) — the point estimate the interval is built around.
- Enter the sample's spread into Sample standard deviation (s) — it cannot be negative.
- Enter how many observations went into that mean into Sample size (n) — larger n narrows the interval.
- Read Lower bound (95% CI) and Upper bound (95% CI) — the true population mean is estimated to lie in that range with 95% confidence.
- Need a different confidence level for the same data? Use this site's 90% or 99% confidence interval instruments, or the adjustable confidence-interval instrument that lets you pick any of five standard levels.
Worked example — sample mean 100, s = 15, n = 25
Enter 100 into Sample mean (x-bar), 15 into Sample standard deviation (s), and 25 into Sample size (n). The instrument first finds the standard error: 15 / √25 = 15 / 5 = 3.0000. It then multiplies that by the fixed 95% critical value: 1.96 × 3.0000 = 5.88 — the margin of error.
Lower bound (95% CI) reads 94.12 (100 − 5.88) and Upper bound (95% CI) reads 105.88 (100 + 5.88). Reported in full: you are 95% confident the true population mean lies between 94.12 and 105.88, the standard way this kind of result is stated in a paper or report.
Questions
Why is 95% the default confidence level almost everywhere?
Not for any deep mathematical reason — it's a convention that stuck, popularised by early 20th-century statisticians and then cemented by decades of journal norms and textbook defaults. It sits at a practical middle ground: tighter and more actionable than 99%, but more trustworthy than 90%. When a report just says 'confidence interval' with no level specified, 95% is the safest assumption.
What does '95% confidence' actually mean about my specific interval?
It describes the reliability of the method, not a probability about this one result. Either the true mean is inside your computed interval or it isn't. What '95%' promises is that if you repeated the sampling and interval-building process many times, about 95% of the resulting intervals would contain the true population mean — 5% would miss it.
Where does the critical value 1.96 come from?
It is the z-value on the standard normal distribution beyond which exactly 2.5% of the area lies in each tail, leaving 95% in the middle — a value published in every standard statistical table and in the NIST/SEMATECH e-Handbook's table of the standard normal distribution. It applies when the normal distribution is an appropriate model, typically for a reasonably large sample.
Should I use 1.96 (normal) or a t-distribution critical value instead?
Use 1.96 when your sample is reasonably large (a common rule of thumb is n ≥ 30) or the population standard deviation is genuinely known. For smaller samples where s is only an estimate of an unknown population sigma, the t-distribution's critical value is more rigorous — it's always somewhat larger than 1.96 and shrinks toward it as sample size grows. See this site's t-statistic instrument for that case.
How do I get a narrower interval without changing the confidence level?
Collect more data. The margin of error shrinks with the square root of n, so quadrupling your sample size only halves the interval's width — a diminishing-returns pattern, not a linear one. Lowering the confidence level (to 90%, say) also narrows the interval, but at the cost of a higher chance of missing the true mean, which is a different trade-off entirely.
Why does this site have separate 90%, 95% and 99% calculators?
Each fixes a different standard critical value so you can jump straight to the conventional level for your field without selecting it manually — 90% for quicker, lower-stakes looks, 95% as the general default, 99% for higher-stakes claims. If you need to compare bounds across levels on the same data, or want a level outside those three, this site's adjustable confidence-interval instrument covers 80% through 99% in one place.