Z-Score Calculator: Standardise Any Value
How many standard deviations from the mean
Try the calculator
Math runs in your browser. No sign-up.
Standard deviations from the mean. Negative means below.
85 is 1.5 standard deviations from a mean of 70.
A z-score is (value β mean) Γ· standard deviation. A score of 85 where the mean is 70 and the standard deviation 10 gives z = 1.5, meaning it sits one and a half standard deviations above average.
A z-score expresses how far a value sits from its mean, measured in standard deviations. It is the fundamental tool for comparing measurements that arrive on different scales β a test mark against a height, a lab result against a rainfall figure β because dividing by the standard deviation strips out the original units entirely. A z of 0 is exactly average, +1 is one standard deviation above, β2 is two below. In a normal distribution, roughly 68% of values fall within Β±1, 95% within Β±2 and 99.7% within Β±3, which is where the common rule of thumb about outliers beyond three standard deviations comes from. The percentile interpretation only holds if the underlying data really is approximately normal.
What is a z-score?
Convert a raw measurement into a z-score β the number of standard deviations it sits from the mean. Standardising strips out the original units, which is what makes values from completely different scales comparable. Enter the value, the mean and the standard deviation to get the score, the raw distance from the mean, and the absolute magnitude for outlier screening. Percentile interpretations assume the underlying data is approximately normal.
The formula
Source: Standard score (z-transformation).
Worked examples
1Comparing a test score against the class
An 85 where the class averaged 70 with a standard deviation of 10 gives z = 1.5. If the marks are roughly normal, that places the student near the 93rd percentile. The same raw 85 in a class averaging 82 with a standard deviation of 3 would give z = 1.0 β a markedly less exceptional result from an identical mark.
2A value below the mean
Sixty against a mean of 70 with a standard deviation of 8 gives z = β1.25, about the 10th percentile under normality. The absolute value of 1.25 is well inside the conventional outlier thresholds of 2 or 3, so this is a low result rather than an anomalous one.
How to use this calculator
- Value β The raw observation you want to standardise.
- Mean (default: 70)
- Standard deviation β Must be greater than zero. Use the population standard deviation where you have the whole population, the sample value otherwise.
- Read the result. Use the worked examples below to sanity-check against a known scenario.
Common mistakes and edge cases
Reading percentiles from non-normal data
The 68β95β99.7 figures are properties of the normal distribution. Applying them to skewed data such as income, waiting times or house prices produces badly wrong percentile estimates even though the z-score itself is still computable.
Confusing population and sample standard deviation
The population form divides by n, the sample form by n β 1. For small datasets the difference is material and inflates or deflates every z-score computed from it.
Treating a negative z as an error
A negative score simply means the value sits below the mean. The sign carries the direction and should not be discarded except when deliberately screening for outlier magnitude.
Using z-scores on tiny samples
With fewer than about 30 observations, the sample standard deviation is itself an unstable estimate. A t-distribution is the appropriate tool in that regime.
Frequently asked questions
how do I calculate a z-score
what is a good z-score
what does a negative z-score mean
at what z-score is a value an outlier
can I use z-scores on data that is not normal
Z-Score glossary
How we built this calculator
Methodology
Standardising works in two steps. Subtracting the mean re-centres the distribution on zero, so the result expresses distance from average rather than an absolute level. Dividing by the standard deviation then rescales that distance into units of typical spread, which removes the original measurement units entirely.
Sources & references
Every numeric assumption traces to a primary source.
- NIST/SEMATECH e-Handbook of Statistical MethodsUSA
- U.S. Census Bureau β Statistical Quality StandardsUSA
- International Organization for Standardization β ISO 3534-1 Statistics VocabularyINT
- Numora Editorial Policy. numora.net/editorial-policy