StatReasonWhat this answers
A z score answers "how many standard deviations away from the reference mean does this specific value sit?" It converts a raw value into a standardized, unit-free number that can be compared across completely different scales, for example comparing a test score to a measurement in a different unit entirely, since both are expressed the same way once standardized.
How it is calculated
The z score is the observed value minus the reference mean, divided by the reference standard deviation. A positive z means the value sits above the reference mean; negative means below. A z score of exactly 0 means the value equals the reference mean exactly.
Worked example
For an observed value of 70, a reference mean of 50, and a reference standard deviation of 10: the z score is (70 minus 50) divided by 10, which is exactly 2, meaning this value sits exactly two standard deviations above the reference mean.
Assumption audit
Common mistakes
Comparing raw values across two different scales directly, without converting either to a z score first, is a common error: a score of 85 on one test and 850 on another cannot be compared meaningfully without knowing each test's own mean and spread. Converting both to z scores against their own reference mean and standard deviation puts them on the same standardized footing, which raw values alone never do.
What this result does not mean
A z score by itself does not tell you a probability or a percentile rank unless you are also willing to assume the underlying data is approximately normally distributed. On clearly skewed data, the same z score corresponds to a different percentile than it would under a normal model.
Limitations
This calculator requires a stated reference mean and standard deviation; it does not estimate those from a sample for you. If you need to compute them from raw data first, use the Mean and Standard Deviation calculators.