📊 Z-Score Calculator

A z-score (also called a standard score) tells you how many standard deviations a particular value sits above or below the population mean. The score turns any normal distribution into the standard normal distribution, which has a mean of zero and a standard deviation of one. That standardization is powerful: it lets you compare values from entirely different scales — a SAT score and an ACT score, a blood pressure reading and a cholesterol reading, or a manufacturing tolerance from two different production lines — on a single, shared yardstick.

Converting a raw score to a z-score is the bridge between a data point and the probability tools built on the bell curve. Once you have the z-value, you can look up the percentile rank (what percentage of the population scores below you), the cumulative probability P(X ≤ x), and the tail probability P(X > x). These are the same quantities produced by a probability calculator and the same distribution modelled by a normal distribution calculator. If you need the spread of the underlying data first, a standard deviation calculator supplies the σ you plug in here.

The concept was introduced by statistician Karl Galton in the late 19th century and formalised through the work of Pearson and Fisher. Today it is taught in every introductory statistics course and used daily in education, psychology, medicine, finance, and Six Sigma quality programmes. A z-score of +1 means the value is one standard deviation above the mean — better than about 84% of the population. A z-score of −2 means the value is two standard deviations below the mean — lower than about 98% of the population.

The z-score is calculated with a single, elegant formula:

where X is the raw score you are testing, μ is the population mean, and σ is the population standard deviation. Subtract the mean to find how far the score sits from centre, then divide by σ to express that distance in standard-deviation units.

Once you have Z, three derived quantities follow directly:

• Percentile rank — the percentage of values in the distribution that fall at or below your score. It equals Φ(Z) × 100, where Φ is the cumulative distribution function of the standard normal. A z of 1.0 gives the 84.1st percentile; a z of 1.96 gives the 97.5th percentile.
• Cumulative probability P(X ≤ x) — the same value expressed as a probability from 0 to 1 instead of a percentage. For z = 1, P(X ≤ x) ≈ 0.841.
• Upper-tail probability P(X > x) — one minus the cumulative probability, giving the chance of seeing a value more extreme than your score. This is the quantity behind p-values in hypothesis testing.

The calculator computes Φ(Z) using the Abramowitz-Stegun approximation of the error function (erf), which is accurate to about 7 decimal places — more than enough for any practical use. For the reverse direction, enter a z-score and the tool recovers the percentile and both tail probabilities without needing μ or σ, because the standard normal distribution is fully defined by Z alone.

1. Decide which direction you need. To go from a raw score to a z-score, fill in the three inputs X (your value), μ (the population mean), and σ (the population standard deviation). To go the other way, enter a z-score directly in the reverse-calculation field.
2. Enter the population mean μ. This is the average of the full group you are comparing against — for example, the class average on an exam or the nominal dimension of a manufactured part.
3. Enter the population standard deviation σ. It must be a positive number. If you have raw data instead, run it through a standard deviation calculator first and copy the population value here.
4. Enter the raw score X and press Calculate. The tool outputs the z-value, the percentile rank, the cumulative probability P(X ≤ x), and a plain-English interpretation such as "above the mean by 1.5 standard deviations."
5. For the reverse direction, type a z-score and press Calculate to recover the percentile, the cumulative probability, and the upper-tail probability — useful for reading p-values or translating standard normal tables.

What does a z-score of 1.5 mean?

A z-score of 1.5 means the value is exactly one and a half standard deviations above the population mean. In a normal distribution, that corresponds to roughly the 93rd percentile — higher than about 93% of all values. It is generally considered a strong result, whether the score is an exam grade, a test statistic, or a process measurement.

Can a z-score be negative?

Yes. A negative z-score means the raw value is below the mean. A z of −1 indicates the score is one standard deviation below average, which falls around the 16th percentile. Negative z-scores are common and perfectly meaningful — they simply show the direction of deviation from centre.

What is the difference between z-score and t-score?

A z-score uses the population standard deviation σ and assumes you know it. A t-score uses the sample standard deviation s and follows the Student t-distribution, which is wider to reflect the extra uncertainty of estimating σ from limited data. For sample sizes above 30 the two are practically identical, so this calculator is a good approximation in most real-world cases.

How is the percentile calculated from the z-score?

The calculator applies the Abramowitz-Stegun approximation of the error function to evaluate the standard normal cumulative distribution Φ(Z). The percentile is Φ(Z) multiplied by 100. The approximation is accurate to roughly seven decimal places, which is far tighter than any z-table you will find in a textbook.

Do z-scores only work for normal distributions?

The z-score formula itself can be computed for any distribution — it is just a linear rescaling. However, the percentile and probability outputs shown here rely on the normal distribution, so they are only trustworthy when the underlying data is approximately bell-shaped. For strongly skewed data, consider a non-parametric percentile or a transformation first.

What z-score corresponds to the 95th percentile?

The 95th percentile of the standard normal distribution is approximately z = 1.645. For a two-tailed 95% confidence interval you would use ±1.96, which leaves 2.5% in each tail. These two numbers — 1.645 and 1.96 — come up constantly in hypothesis testing and sample size planning.

A positive z-score means the raw value is above the mean; a negative one means it is below. The sign matters — a z of −1.5 is not the same as +1.5, even though both are "one and a half standard deviations" from centre.

Z-scores only make sense when the underlying distribution is roughly normal. If your data is heavily skewed or bimodal, the percentile derived from the bell curve will be misleading. Always sanity-check the distribution shape first, and if you do not yet know σ, compute it with a standard deviation calculator.

Watch out for sample versus population standard deviation. This calculator expects the population σ. If you only have a sample, the z-score is technically an approximation; the more precise alternative is a t-score, but for n above 30 the difference is negligible.

The empirical rule (68–95–99.7) is a handy sanity check: about 68% of values fall within ±1σ, 95% within ±2σ, and 99.7% within ±3σ. If your percentile result contradicts those landmarks, recheck your inputs.

Z-scores are dimensionless. You can compare a temperature reading in Celsius with a test score out of 100 — both become the same unit-free number once standardised. That is why z-scores underpin standardised exams, growth charts, and process capability indices alike.