Z-Score Calculator
Calculate z-score, percentile rank, and probability from a normal distribution.
PhD Statistics, Fellow of the Royal Statistical Society
Statistician and data scientist with 15 years in applied statistics, probability theory and data visualisation across industry and academia.
Related calculators
About the Z-Score Calculator
The z-score (also called a standard score) is one of the most powerful ideas in statistics: it transforms any raw measurement into a universal currency of "how unusual is this value?" by expressing it as a number of standard deviations above or below the population mean. A z-score of 0 means the value is exactly average; +1 means one standard deviation above average; โ2 means two standard deviations below. Because the normal distribution is symmetric and well-characterised mathematically, any z-score can be converted to a percentile โ telling you what percentage of the population falls below that value.
Z-scores are used across diverse domains: IQ tests are normalised so that mean IQ = 100 (z=0) and standard deviation = 15 (so IQ 115 = z=+1, IQ 130 = z=+2, IQ 145 = z=+3). SAT scores (mean 1060, SD 217) work identically โ a score of 1277 is one SD above average (roughly 84th percentile). In medicine, bone density T-scores are z-scores comparing an individual to a young adult reference population; Z-scores compare to age-matched peers. Blood test results are often flagged as abnormal based on whether they fall outside 1.96 SDs (the 95th percentile boundary).
The conversion from z-score to percentile uses the cumulative distribution function (CDF) of the normal distribution, which has no closed-form expression and must be computed numerically โ this calculator uses the Abramowitz & Stegun approximation, accurate to within 0.00001. A key table to memorise: z=ยฑ1 covers 68.27% of the distribution; z=ยฑ2 covers 95.45%; z=ยฑ3 covers 99.73%. This is the "68-95-99.7 rule" (or empirical rule) that appears throughout statistics, quality control, and scientific measurement.
How it works
Z = (x โ ฮผ) / ฯ Percentile = ฮฆ(Z) ร 100 [where ฮฆ is the normal CDF]
Where
xThe observed value you want to standardiseฮผPopulation mean (the average value in the reference population)ฯPopulation standard deviation (spread of the reference population)ZZ-score โ number of standard deviations x is from the meanฮฆ(Z)Normal CDF โ the area under the standard normal curve to the left of ZWorked example
Example: A student scores 82 on an exam. The class mean is 70, standard deviation is 10.
Z = (82 โ 70) / 10 = 1.2.
Using the normal CDF: ฮฆ(1.2) = 0.8849 โ 88.5th percentile.
Interpretation: The student scored better than approximately 88.5% of the class.
Example 2: A man is 185cm tall. UK male mean height โ 175.3cm, SD โ 7.2cm.
Z = (185 โ 175.3) / 7.2 = 1.347.
ฮฆ(1.347) = 0.9109 โ approximately 91st percentile for UK adult males.
Tips to improve your result
- 1.
The 68-95-99.7 rule is invaluable to memorise: 68% of data falls within ยฑ1 SD; 95% within ยฑ2 SD; 99.7% within ยฑ3 SD. In quality control (Six Sigma), "6 sigma" quality means defect rates within ยฑ6 SD โ approximately 3.4 defects per million opportunities.
- 2.
Z-scores require that the data is approximately normally distributed. Many real-world distributions are not: income is log-normal (right-skewed), biological lifespans have complex shapes, and extreme events (financial crashes, earthquakes) follow power laws. Applying z-score interpretation to non-normal data gives misleading percentile rankings.
- 3.
When working with sample data rather than population data, use the t-distribution instead of the normal distribution if n < 30. The t-distribution has "heavier tails" that account for the extra uncertainty from estimating both the mean and SD from a small sample. As n increases, the t-distribution approaches the normal distribution.
- 4.
In psychological testing, scores are often reported as "standard scores" (mean 100, SD 15) or T-scores (mean 50, SD 10) rather than z-scores. These are simply linear transformations of the z-score: Standard score = z ร 15 + 100; T-score = z ร 10 + 50. The underlying z-score is the same concept.
- 5.
A z-score of ยฑ1.96 corresponds to the 97.5th and 2.5th percentiles โ together these define the 95% central region, which is why 1.96 appears so often in confidence interval calculations (ยฑ1.96 ร SE gives a 95% confidence interval).