What is a confidence interval?

A confidence interval is a range of values calculated from sample data that is likely to contain the true population parameter. A 95% confidence interval for a mean, for example, provides a lower and upper bound such that 95% of similarly constructed intervals would contain the true population mean. It is not a statement that there is a 95% probability the true mean is in this specific interval — the true mean is fixed; the interval is what varies across samples.

How do you calculate a 95% confidence interval?

For a population mean: CI = x̄ ± z × (s / √n), where x̄ is the sample mean, s is the sample standard deviation, n is the sample size, and z is the critical z-value. For 95% confidence, z = 1.96. For example, if x̄ = 100, s = 15, and n = 36, the standard error is 15/√36 = 2.5 and the margin of error is 1.96 × 2.5 = 4.9. The 95% CI is [95.1, 104.9].

What is the margin of error?

The margin of error is half the width of the confidence interval — it equals z × (s / √n). It represents the maximum expected difference between the sample estimate and the true population value. A larger margin of error means less precision. In news polling, a "±3 percentage points" margin of error means the confidence interval extends 3 points above and below the reported percentage. You can reduce the margin of error by increasing the sample size or accepting a lower confidence level.

What does "95% confident" actually mean?

If you repeated your study many times — each time drawing a new sample and calculating a confidence interval — 95% of those intervals would contain the true population mean. The other 5% would miss it. For any single study, the interval either contains the true mean or it does not. The 95% refers to the long-run reliability of the method, not the probability for a single result. Researchers choose 95% as a balance between precision and reliability — it is the most common standard in science.

How does sample size affect the confidence interval?

Increasing sample size reduces the confidence interval width. The margin of error is proportional to 1/√n, so quadrupling the sample size halves the margin of error. For example, with s = 10 and 95% confidence: n = 25 gives ME = 1.96 × 10/5 = 3.92; n = 100 gives ME = 1.96 × 10/10 = 1.96; n = 400 gives ME = 1.96 × 10/20 = 0.98. This is why large clinical trials and surveys are more precise than small ones — the larger n narrows the confidence interval.

What is the difference between a confidence interval and a prediction interval?

A confidence interval estimates where the population mean lies. A prediction interval estimates where a future individual observation will fall. Prediction intervals are always wider than confidence intervals because they must account for both the uncertainty in estimating the mean and the natural variability of individual observations. For example, a 95% CI for average height might be [175, 177] cm, while a 95% prediction interval for one person's height might be [155, 197] cm — much wider because individuals vary more than means.

How do I reduce the margin of error in my confidence interval?

The margin of error equals z × (s / √n), so you can shrink it by: (1) increasing sample size n — quadrupling n halves the margin of error; (2) lowering the confidence level from 99% to 95% or 90%, which uses a smaller z-value; or (3) reducing variability in your measurement process (lower s). In practice, increasing n is the most reliable approach, though it requires more data collection time and cost.

Where are confidence intervals used in real life?

Confidence intervals appear in clinical trials (a drug reduces blood pressure by 5–9 mmHg with 95% confidence), political polling (candidate leads by 4 ± 3 percentage points), quality control (average widget weight is 10.2 ± 0.1 grams), and scientific research publications (nearly every study reports CIs alongside p-values). The 2020 shift in medical journals away from p-values toward CIs reflects that CIs convey both statistical significance and practical magnitude.

Confidence Interval Calculator — Mean & Margin of Error

Calculates the confidence interval, margin of error, and standard error for a population mean at 90%, 95%, or 99% confidence.

How to Use the Confidence Interval Calculator

This confidence interval calculator finds the interval, margin of error, and standard error for any population mean — select your desired confidence level (90%, 95%, or 99%), then enter the sample mean (x̄), sample standard deviation (s), and sample size (n). The calculator instantly displays the confidence interval bounds, margin of error, and standard error. For a custom z-value — for example, for 99.9% confidence — select "Custom z" and enter the critical value directly.

The output includes a plain-language interpretation: "We are 95% confident the true population mean falls between [lower] and [upper]." For questions about the spread of your data before calculating a confidence interval, use our standard deviation calculator.

The Confidence Interval Formula

For a population mean, the confidence interval is built from the sample mean, the critical z-value for the chosen confidence level, and the standard error:

CI = x̄ ± z × (s / √n)

x̄ — sample mean
z — critical z-value: 1.645 (90%), 1.96 (95%), 2.576 (99%)
s — sample standard deviation
n — sample size
s / √n — standard error of the mean (SEM)

Worked Example

A researcher measures blood pressure in 50 patients: x̄ = 120 mmHg, s = 10 mmHg. Find the 95% confidence interval.

Standard error: SE = 10 / √50 = 10 / 7.071 = 1.414
Margin of error: E = 1.96 × 1.414 = 2.77
Lower bound: 120 − 2.77 = 117.23
Upper bound: 120 + 2.77 = 122.77
95% CI: [117.23, 122.77]

Interpretation: We are 95% confident the true mean blood pressure in the population lies between 117.23 and 122.77 mmHg.

AdvertisementResponsive Ad

Choosing the Right Confidence Level

The three most common confidence levels each represent a trade-off between precision and reliability:

90% (z = 1.645): Narrower interval, less certain. Used when a tighter estimate is more useful than high reliability — for example, in internal business analyses where decisions can be adjusted quickly.
95% (z = 1.96): The scientific and statistical default. Required by most academic journals, clinical trials, and public health studies. Balances precision and reliability.
99% (z = 2.576): Wider interval, very reliable. Used when the cost of being wrong is high — for example, in pharmaceutical drug approvals, engineering safety margins, and financial stress tests.

Increasing confidence level always widens the interval. There is no free lunch: higher confidence means you must accept less precision (a wider range of plausible values) for the same sample size.

Standard Error vs. Standard Deviation

The confidence interval uses the standard error (SE = s / √n), not the standard deviation. These measure different things:

Standard deviation (s): Measures how much individual data points vary from the sample mean. It describes the spread of the data itself.
Standard error (SE = s / √n): Measures how much the sample mean would vary if you repeated the study with different samples. It describes the precision of the mean estimate, not the spread of the data.

Standard error is always smaller than standard deviation (because SE = s / √n and √n > 1 for n > 1). As n increases, SE decreases, which narrows the confidence interval. For more detail on standard error, see our standard error calculator.

AdvertisementResponsive Ad

Sample Size Planning

You can work backwards from a desired margin of error to determine the required sample size. Rearranging the formula:

n = (z × s / E)²

Where E is the desired margin of error. For example, to achieve a margin of error of ±2 with s = 10 at 95% confidence:

n = (1.96 × 10 / 2)² = 9.8² = 96.04 → round up to n = 97

Always round up (never down) when calculating required sample size. In survey research, an additional 10–20% buffer is common to account for non-response and data quality issues.

Real-World Uses of Confidence Intervals

Confidence intervals are ubiquitous in science, medicine, business, and public policy. In clinical trials, a drug is approved when its confidence interval for the treatment effect excludes zero — for example, “reduces systolic blood pressure by 4 to 8 mmHg (95% CI)” clearly shows a real benefit. Political polls report margins of error (typically ±3 percentage points for national samples of ~1,000) to convey polling uncertainty. In A/B testing for websites, engineers use CIs to decide whether a new feature lifts conversion rates enough to justify a rollout. Quality engineers use CIs on manufacturing measurements to confirm that products meet tolerance specifications. Virtually every peer-reviewed scientific result now reports a confidence interval alongside the p-value — because a CI conveys both statistical significance and the practical magnitude of the effect.

Sources & References

Confidence Intervals — NIST/SEMATECH e-Handbook of Statistical Methods
Confidence Intervals — Khan Academy — Khan Academy

Confidence Interval Calculator