Standard error tells you how much a sample mean is likely to differ from the true population mean, just from the randomness of sampling. It's one of the most-used — and most-confused — statistics in research, because it's easy to mix up with standard deviation. This guide explains what standard error measures, how it's calculated, and where it's used.
Standard error vs. standard deviation
Standard deviation (SD) describes how spread out individual data points are around the mean — it's a property of your raw data. Standard error (SE) describes how much the sample mean itself would vary if you repeated the sampling process many times — it's a property of your estimate, not your data. The relationship is simple: SE = SD / √n. As your sample size n grows, the standard deviation of the data stays roughly the same, but the standard error of the mean shrinks, because larger samples give more stable, precise estimates of the true average.
Standard error is useful for describing estimated sampling variability, but a confidence interval requires multiplying it by an appropriate critical value, often a t critical value for a small sample. Standard deviation remains the right measure when describing the spread of individual observations — for example, in error bars on a chart of average measurements, or in the denominator of a t-test statistic.
Why the square root of n matters
Standard error doesn't shrink in direct proportion to sample size — it shrinks with the square root of n. Doubling your sample size only reduces the standard error by a factor of about 1.41 (√2), not by half. To cut the standard error in half, you need four times the sample size; to cut it to a third, you need nine times the sample. This diminishing-returns relationship is why very large sample sizes are needed to squeeze out small improvements in precision, and why researchers weigh the cost of additional data collection against how much it will actually tighten their estimate.
Limits and what standard error doesn't tell you
Standard error only accounts for random sampling variability — it assumes your sample was drawn independently and randomly from the population you care about. It says nothing about measurement bias, a poorly designed sample, or systematic errors in data collection; a study can report a tiny standard error and still be badly wrong if the underlying sample isn't representative. Standard error is also the building block for other statistics — it's multiplied by a critical value (a z-score or t-score) to build a margin of error and a confidence interval, and it appears in the denominator of hypothesis-testing statistics like the t-statistic.