Evidence-appraisal glossary
Standard error
Standard error estimates how much a summary statistic, such as a sample mean, would vary from one sample to another; it reflects the precision of the estimate, not the spread of the data.
Also called: SE, SEM.
The standard error of a mean equals the standard deviation divided by the square root of the sample size, so it gets smaller as the study enrolls more people, signaling a more precise estimate. Confidence intervals are built directly from it. Do not read it as the typical distance of individual values from the mean, which is what the standard deviation describes.
Read the full Reading the Evidence blog.
This is a plain-language methodology definition for reading research. It is general education, not medical advice.