What This Calculator Does
This standard error calculator computes SE for both the sample mean and the sample proportion, then constructs a confidence interval at the chosen confidence level. Standard error quantifies sampling variability, so smaller SE means a more reliable estimate. The calculator is designed for researchers, students, analysts, and survey designers who need a quick, accurate SE without opening a statistics package.
Standard Error Formulas
For the Sample Mean
SE = σ / √n
- σ: standard deviation (population or sample estimate)
- n: sample size
For the Sample Proportion
SE = √(p(1 − p) / n)
- p: sample proportion (decimal form, e.g. 0.5 not 50)
- n: sample size
Confidence Interval
A confidence interval is estimate ± z × SE. The z critical value depends on the chosen confidence level: 1.645 for 90%, 1.96 for 95%, 2.576 for 99%. For small samples (n < 30) with unknown σ, use the t-distribution instead.
Worked Example (Mean)
Suppose a sample of n = 64 produces a mean of 100 with standard deviation 15. Then:
- SE = 15 / √64 = 15 / 8 = 1.875
- 95% margin = 1.96 × 1.875 ≈ 3.675
- 95% CI = 100 ± 3.675 = (96.325, 103.675)
Worked Example (Proportion)
A poll of n = 1,000 voters finds 52% support for a candidate. Then:
- SE = √(0.52 × 0.48 / 1000) = √0.0002496 ≈ 0.0158
- 95% margin = 1.96 × 0.0158 ≈ 0.0310 (3.1 percentage points)
- 95% CI for support = 52% ± 3.1% = (48.9%, 55.1%)
How to Interpret Standard Error
- Precision: smaller SE = more precise estimate.
- Reliability: a 95% CI has roughly 95% coverage probability in repeated sampling.
- Comparisons: when comparing two groups, smaller SEs make differences easier to detect.
- Sample size planning: solve n = (z × σ / target SE)² to find the required n.
Limitations
SE assumes simple random sampling. For clustered, stratified, or weighted samples, design effects (DEFF) inflate the variance and SE. The formulas here also assume approximate normality, which is reasonable for means when n ≥ 30 and for proportions when np and n(1−p) are at least 5. For non-normal or skewed data, use bootstrap or non-parametric methods.
Sources and References
- Cochran, W. G. Sampling Techniques.
- NIST/SEMATECH e-Handbook of Statistical Methods.
- Lohr, S. Sampling: Design and Analysis.
- Casella, G. & Berger, R. L. Statistical Inference.