Express a data value as a signed number of standard deviations from a mean. Formula and z-score remain visible in one place for independent verification.
A z-score standardizes a value by subtracting the mean and dividing by the standard deviation. Positive scores lie above the mean, negative scores lie below it, and zero sits exactly at the mean.
In z-score, the relationship among the quantities matters just as much as their arithmetic.
Compute the signed difference between the value and mean, then divide by a positive standard deviation. Preserve the sign because it identifies which side of the mean contains the observation.
A score of 78 in a distribution with mean 70 and standard deviation 4 has z = (78 − 70)/4 = 2. It lies two standard deviations above the mean. This Z-Score example can be compared with endpoint-based center.
Standardization puts differently scaled measurements onto a comparable axis. It helps compare test results, spot unusually distant observations, and prepare variables for statistical analysis.
A z-score alone does not prove that a value is rare unless the distribution shape and reference population are understood. Population and sample standard deviations can also produce slightly different scores. If the Z-Score assumptions do not fit, consider absolute spread.
An impossible z-score sign or magnitude usually points to field assignment before it points to rounding.
The input labels—data value, mean and standard deviation—encode the model used on this page. Write those labels beside source values when transferring a problem from paper or a spreadsheet. That small step catches transposed quantities and mixed units before they become a polished-looking z-score.
Raising the data value by one standard deviation raises z by one. Increasing the standard deviation with value and mean fixed pulls the score toward zero. A one-field trial makes this relationship visible without reworking the entire example.
Choose rounding after considering how the result will be used. Comparison may need only a few significant digits, while a later multi-step calculation benefits from carrying more. In either case, retain the page's formula with the z-score so the underlying definition remains visible.
A z-score describes standardized distance, not probability by itself. Converting it to a tail area requires a distribution assumption that this calculator does not silently impose. The distinction determines whether this page fits the original question.
If the number moves into a spreadsheet, give its cell a z-score heading and retain the source values data value, mean and standard deviation nearby. Context matters more than extra displayed digits.
The observation equals the stated mean.
Yes. It simply indicates a value more than three standard deviations from the mean.
A zero spread provides no scale for division, and a standard deviation is never negative.
Only after adopting a distribution model, commonly the normal distribution.