THE NUMORIX GUIDE
How to use the Z-Score Calculator
Last reviewed September 14, 2026
What this calculator does
Calculate z = (value - mean)/standard deviation.
Formula and method
Calculate z = (value - mean)/standard deviation. The sign shows whether the value is above or below the mean, and the magnitude counts standard deviations.
Variables and inputs
Value defaults to 100, Mean to 90, and Std Dev to 15. Standard deviation must be nonzero; it should ordinarily be positive.
Worked example
For value 100, mean 90, and SD 15: z = (100-90)/15 = 10/15 = 0.6667. The value is about two-thirds of a standard deviation above the mean.
How to interpret the result
A z-score of 0 is exactly average; positive values are above average and negative values below average, under the selected mean and spread.
Common mistakes to avoid
Do not divide by variance instead of standard deviation. Use the standard deviation from the same population or sample definition as the mean.
Assumptions and limitations
A z-score standardizes location but does not prove a normal distribution or provide a percentile by itself. A zero SD makes the calculation undefined.
Practical use and checks
A z-score places one observation relative to a mean in standard-deviation units. Enter value 100, mean 90, and standard deviation 15 as a check; the result should be 0.6667, meaning the value is about two-thirds of a standard deviation above the mean. Entering value 75 with the same mean and spread should return -1, which indicates one standard deviation below the mean. Use the sign for direction and the magnitude for distance. The mean and standard deviation must describe the same population, sample definition, and measurement units as the value. A z-score is useful for comparing differently scaled tests or identifying observations that deserve review, but it is not automatically a percentile. Under a standard normal model, z = 2 has a familiar tail interpretation; in a skewed or small dataset, that translation may be wrong. A zero standard deviation makes the denominator undefined, and a negative standard deviation is not a meaningful spread input even if a raw engine call accepts it. The calculator does not test normality, outliers, or whether a reference mean is stable. Record the source of the mean and spread before using the score to label a person, process, or measurement as unusual.