Z Score Calculator

A raw score means little on its own — 86 on one test might be ordinary and on another exceptional. The z-score calculator solves that by restating any value in standard-deviation units: it tells you exactly how far above or below the mean a result sits. Teachers use it to compare scores across different exams, researchers use it to spot unusual observations, and quality teams use it to flag measurements that drift too far from the process average.

Quick Answer

z = (x − μ) ÷ σ, where x is your value, μ is the mean of the distribution, and σ is its standard deviation. A score of 86 in a class with mean 70 and standard deviation 8 gives z = 16 ÷ 8 = 2.0 — two standard deviations above average.

Reading a Z-Score Like a Statistician

The sign of a z-score carries the direction: positive means above the mean, negative means below it, and zero means the value is the mean. The size carries the rarity. Thanks to the empirical rule for roughly bell-shaped data, about 68% of values fall within z = −1 to +1, about 95% within −2 to +2, and about 99.7% within −3 to +3. That is why a z-score beyond ±3 raises eyebrows in almost any field — under a normal distribution it should happen only about three times in a thousand.

Z-scores also convert directly into percentiles when the data is approximately normal. A z of 1.0 sits near the 84th percentile, meaning about 84% of values fall below it. A z of 2.0 reaches about the 97.7th percentile. Going the other way, z = −1.5 leaves only about 6.7% of values below it, near the 7th percentile. This translation is the z-score's real power: it turns "my score was 86" into "my score beat roughly 98% of the group," a statement that stays meaningful even when the next test is marked out of 50 instead of 100.

Two ingredients must come from the same dataset as the value itself. The mean and standard deviation of your class describe your class; plugging in a national average with your school's standard deviation produces a number that describes nothing real.

How to Use the Z-Score Calculator

1. Gather the three numbers from the same dataset: the value you want to evaluate, the mean, and the standard deviation.
2. Enter the individual value (x) first.
3. Enter the mean (μ) of the distribution it came from.
4. Enter the standard deviation (σ), making sure it is not zero — a zero spread makes the z-score undefined.
5. Read the z-score and note its sign: positive is above average, negative is below.
6. If you need a percentile, check that the data is roughly bell-shaped before translating, since the conversion assumes normality.

One Exam, Two Students, Two Very Different Stories

A class takes an exam with a mean score of 70 and a standard deviation of 8. Student A scores 86. Her z-score is (86 − 70) ÷ 8 = 16 ÷ 8 = 2.0, placing her around the 97.7th percentile — a genuinely top-of-class result. Student B scores 58. His z-score is (58 − 70) ÷ 8 = −12 ÷ 8 = −1.5, around the 7th percentile. The raw gap between them is 28 points, but the z-scores say something the raw numbers cannot: A's result is rarer than B's is weak, because the distribution is not symmetric around them in practical terms — B is 1.5 standard deviations out, A is a full 2.

Reference points worth memorising:

• z = 0: exactly average, 50th percentile
• z = +1: about the 84th percentile
• z = +2: about the 97.7th percentile
• z = −2: about the 2.3rd percentile

Where Z-Score Calculations Go Wrong

The subtlest error involves which standard deviation you feed in. Statistics distinguishes the population standard deviation (σ, computed over every member) from the sample standard deviation (s, estimated from a subset, dividing by n − 1). For a z-score describing a member of a fully known group, use the population figure; mixing the two quietly shifts every result. A second trap is forcing the percentile interpretation onto skewed data. Income, reaction times and website load times are lopsided, so a z of 2 there does not mean the 97.7th percentile — the z-score itself is still computable, but the percentile translation is not valid.

Sign errors are the everyday blunder: subtract in the wrong order and an above-average student appears below average. And remember what a z-score cannot do — it does not fix a biased measurement or a tiny sample. With only five observations, the mean and standard deviation are shaky estimates, and every z built on them inherits that shakiness.

Standard Scores Beyond the Classroom

The same standardisation appears in disguise across many fields. IQ scores are z-scores in costume: the tests are scaled to a mean of 100 and a standard deviation of 15, so an IQ of 130 is exactly z = (130 − 100) ÷ 15 = 2.0 — the same rarity as Student A's exam result. Paediatric growth charts work identically, expressing a child's height as a position within the distribution for their age, which is why a doctor can say a toddler tracks along the 90th percentile line year after year even as the raw centimetres change completely. In finance, returns are compared across assets with different volatilities by dividing by their standard deviations, a z-score in all but name.

The formula also runs backwards. Given a mean of 70, a standard deviation of 8, and a target z of 1.5, the required raw score is x = μ + zσ = 70 + 1.5 × 8 = 82. Teachers use this direction to set grade boundaries: decide first what share of the class should earn each grade, translate those shares into z values, and the cut-off scores fall out. Students can use the same trick in reverse — knowing the class statistics and the boundary, you can work out exactly what a target grade demands before you sit the paper.

From Z-Scores to the Rest of Your Analysis

The z-score is usually a middle step, not a final answer. It feeds probability questions (what share of values fall between two scores?), outlier screening, and standardised comparisons across totally different scales, from test results to manufacturing tolerances. If you need the spread figures first, the standard deviation calculator and the mean calculator will prepare them, and the probability calculator helps with the follow-up questions. More statistics tools live in the math calculators section.