Z-Score Calculator

What Is a Z-Score?

A z-score (also called a standard score or z-value) measures how many standard deviations a data point is from the mean of its distribution. The formula is z = (x - mu) / sigma, where x is the data point, mu is the population mean, and sigma is the population standard deviation. A z-score of 0 means the value equals the mean, a positive z-score means the value is above the mean, and a negative z-score means the value is below the mean. For example, if exam scores have a mean of 75 and standard deviation of 10, a score of 85 has a z-score of (85 - 75) / 10 = 1.0, meaning it is exactly one standard deviation above the mean.

This calculator computes the z-score from the raw value, mean, and standard deviation you provide, then converts it to a percentile and cumulative probability. The z-score standardizes values from any normal distribution to the standard normal distribution (mean = 0, standard deviation = 1), making it possible to compare values from completely different datasets. A student's height relative to classmates and their test score relative to the class can both be expressed as z-scores, enabling meaningful comparison despite having different units and scales.

The Standard Normal Distribution

The standard normal distribution is a bell-shaped probability distribution with mean 0 and standard deviation 1. Any normal distribution can be converted to the standard normal by computing z-scores. The probability density function is phi(z) = (1 / sqrt(2*pi)) * e^(-z^2/2), and the cumulative distribution function Phi(z) = P(Z <= z) gives the probability that a standard normal random variable is less than or equal to z. This CDF is the function used to convert z-scores to percentiles.

The standard normal distribution has several key properties. It is symmetric about z = 0, so P(Z > z) = P(Z < -z). The total area under the curve equals 1. About 68.27% of values fall within one standard deviation of the mean (z between -1 and 1), about 95.45% fall within two standard deviations (z between -2 and 2), and about 99.73% fall within three standard deviations (z between -3 and 3). This is known as the 68-95-99.7 rule (or empirical rule) and provides a quick way to assess how unusual a particular z-score is.

How to Calculate a Z-Score Step by Step

Computing a z-score requires three pieces of information: the individual value (x), the mean of the population or sample (mu), and the standard deviation (sigma). Step 1: Subtract the mean from the individual value to get the deviation: x - mu. Step 2: Divide the deviation by the standard deviation: (x - mu) / sigma. The result is the z-score. For example, if a factory produces bolts with a mean diameter of 10mm and standard deviation of 0.1mm, a bolt measuring 10.25mm has z = (10.25 - 10) / 0.1 = 2.5, meaning its diameter is 2.5 standard deviations above the mean.

When working with samples rather than populations, use the sample mean (x-bar) and sample standard deviation (s) instead of mu and sigma. If you are computing the z-score for a sample mean (rather than an individual observation), use the standard error of the mean: z = (x-bar - mu) / (sigma / sqrt(n)), where n is the sample size. This formula accounts for the fact that sample means have less variability than individual observations, and it is the basis of z-tests in hypothesis testing.

Z-Score to Percentile Conversion

The percentile corresponding to a z-score tells you what percentage of values in the distribution fall below that point. A z-score of 0 corresponds to the 50th percentile (the median), because half the values are below the mean. Common conversions: z = -2.33 is the 1st percentile, z = -1.645 is the 5th percentile, z = -1.28 is the 10th percentile, z = -0.674 is the 25th percentile, z = 0.674 is the 75th percentile, z = 1.28 is the 90th percentile, z = 1.645 is the 95th percentile, and z = 2.33 is the 99th percentile.

This calculator uses a highly accurate polynomial approximation to compute the cumulative normal probability. The Abramowitz and Stegun approximation used here achieves precision to several decimal places, which is sufficient for virtually all practical applications. For z-scores beyond +/- 4, the percentile approaches 100% or 0% so closely that the difference is negligible. To find the probability between two z-scores, compute P(z1 < Z < z2) = Phi(z2) - Phi(z1), where Phi is the CDF.

The Z-Score Table (Standard Normal Table)

Before calculators and computers, z-score tables (also called z-tables or standard normal tables) were essential tools for statisticians. These tables list the cumulative probability P(Z <= z) for z-values typically ranging from -3.49 to 3.49 in increments of 0.01. To use the table, find the row for the first decimal place and the column for the second decimal place. For example, to find P(Z <= 1.73), look at row 1.7 and column 0.03 to read 0.9582, meaning 95.82% of values fall below z = 1.73.

Key z-values for common significance levels: z = 1.645 (one-tailed 5% test), z = 1.96 (two-tailed 5% test or one-tailed 2.5%), z = 2.326 (one-tailed 1% test), z = 2.576 (two-tailed 1% test). In the two-tailed context, the critical z-value of 1.96 means that if the test statistic exceeds 1.96 or falls below -1.96, you reject the null hypothesis at the 5% significance level. This calculator makes table lookups unnecessary by computing exact probabilities for any z-score.

Z-Scores in Hypothesis Testing

The z-test is a hypothesis test used when the population standard deviation is known and the sample size is large enough for the Central Limit Theorem to apply (typically n >= 30). The test statistic is z = (x-bar - mu_0) / (sigma / sqrt(n)), where x-bar is the sample mean, mu_0 is the hypothesized population mean, sigma is the known population standard deviation, and n is the sample size. If this z-value falls in the rejection region (beyond the critical value), you reject the null hypothesis.

For a one-tailed test at the 5% significance level, the critical z-value is 1.645 (or -1.645 for a left-tailed test). For a two-tailed test at the 5% level, the critical values are +/- 1.96. The p-value is the probability of observing a test statistic as extreme as (or more extreme than) the one computed, assuming the null hypothesis is true. If the p-value is less than the significance level alpha, reject the null hypothesis. For example, if z = 2.15 in a two-tailed test, p = 2 * P(Z > 2.15) = 2 * 0.0158 = 0.0316, which is less than 0.05, so you reject H0 at the 5% level.

Applications of Z-Scores

Z-scores are used across many fields. In education, standardized test scores like the SAT and GRE are reported as scaled scores derived from z-scores, allowing comparison across different test administrations. In finance, z-scores quantify how far a stock's return is from its historical average, and the Altman Z-score predicts the probability of corporate bankruptcy. In medicine, growth charts use z-scores to assess whether a child's height or weight is typical for their age and sex -- a z-score below -2 may indicate underweight or stunting.

In quality control, z-scores determine whether a manufactured product falls within acceptable tolerance limits. The Six Sigma methodology gets its name from the goal of keeping z-scores within +/- 6 standard deviations, corresponding to a defect rate of 3.4 per million opportunities. In weather and climate science, z-scores express how unusual a temperature or precipitation value is compared to historical averages. In sports analytics, z-scores normalize player statistics across different eras, leagues, or positions, enabling fair comparisons. Any situation where you need to compare a value to a distribution benefits from z-score analysis.

Z-Score vs. T-Score

The z-test assumes the population standard deviation is known, but in practice it often is not. When sigma is unknown and must be estimated from the sample, the t-test is used instead, which accounts for the additional uncertainty by using the t-distribution rather than the standard normal distribution. The t-distribution has heavier tails than the normal distribution, meaning it assigns more probability to extreme values, which makes the test more conservative.

The t-distribution is parameterized by degrees of freedom (df = n - 1 for a one-sample test). As sample size increases, the t-distribution approaches the standard normal distribution, and by n = 30 or so, the two are nearly identical. For very large samples, the z-test and t-test give virtually the same results. The general rule is: use a z-test when sigma is known and the sample is large, use a t-test when sigma is unknown (which is most real-world situations). Many introductory statistics courses start with the z-test because it is simpler, then introduce the t-test as the more practical tool.

Limitations and Assumptions

Z-scores and the associated probabilities assume the data follows a normal distribution. For data that is skewed, multimodal, or otherwise non-normal, the probability interpretations may be inaccurate. The z-score itself can always be calculated (it is just arithmetic), but the conversion to percentile using the standard normal CDF is only valid under the normality assumption. Tests for normality include the Shapiro-Wilk test, the Anderson-Darling test, and visual inspection using histograms or Q-Q plots.

The Central Limit Theorem provides some protection: even if the underlying data is not normal, the distribution of sample means approaches normality as sample size increases, typically by n = 30. This means z-tests on sample means are robust to non-normality for large samples, even if the individual observations are not normally distributed. However, for individual observations from non-normal distributions, non-parametric alternatives like percentile ranks (which do not assume normality) may be more appropriate than z-scores.