A z-score, also known as a standard score, quantifies exactly how far a particular data point lies from the mean of its distribution, measured in units of standard deviation. The formula is straightforward: z = (x − μ) / σ, where x is the observed value, μ (mu) is the population mean, and σ (sigma) is the population standard deviation. A z-score of zero means the value is exactly at the mean; positive z-scores indicate values above the mean, and negative z-scores indicate values below it.

The Standard Normal Distribution

When you convert raw data into z-scores, you transform the original distribution into the standard normal distribution, which has a mean of 0 and a standard deviation of 1. This bell-shaped curve is symmetric about the mean, and its total area under the curve equals 1 (or 100%). The beauty of standardization is that no matter what the original units were (test scores, heights, temperatures), every z-score lives on the same universal scale. This makes it possible to compare values from completely different distributions on equal footing.

The 68-95-99.7 Rule (Empirical Rule)

For any normal distribution, approximately 68% of values fall within one standard deviation of the mean (z-scores between −1 and +1), about 95% fall within two standard deviations (z-scores between −2 and +2), and roughly 99.7% fall within three standard deviations (z-scores between −3 and +3). This rule provides a quick mental framework for judging how unusual a particular z-score is. Any z-score beyond ±2 is considered uncommon, and beyond ±3 is rare.

Z-Scores and the Z-Table

A z-table (also called the standard normal table) provides the cumulative probability for each z-score, that is, the probability that a randomly selected value from the distribution is less than or equal to that z-score. For example, a z-score of 1.96 corresponds to a cumulative probability of 0.975, meaning 97.5% of values lie below it. Z-tables are organized with z-scores to one decimal place in the rows and the second decimal place in the columns. While modern calculators and software have largely replaced printed z-tables, understanding how to read one builds statistical intuition.

Z-Scores in Grading and Assessment

Educators use z-scores to grade on a curve by converting raw exam scores into standardized scores. A student who earns a z-score of +1.5 performed 1.5 standard deviations above the class average, regardless of whether the exam was easy or difficult. This approach fairly accounts for exam difficulty and score variability. Z-scores also power the T-score scale used in standardized tests (T = 10z + 50), the SAT's scaled scoring, and IQ scores (IQ = 15z + 100).

Relationship to Percentiles

Every z-score maps to a specific percentile in the normal distribution. A z-score of 0 corresponds to the 50th percentile (median), a z-score of 1.0 to the 84.1st percentile, and a z-score of −1.0 to the 15.9th percentile. Converting between z-scores and percentiles is essential for interpreting standardized test results, quality control measurements, and medical reference ranges.

One-Tailed vs. Two-Tailed Tests

In hypothesis testing, a one-tailed test examines whether a value is significantly greater than (or less than) a threshold, using only one end of the distribution. A two-tailed test checks for significance in either direction, splitting the significance level across both tails. At the common 5% significance level, the one-tailed critical z-score is 1.645, while the two-tailed critical z-score is 1.96. Choosing the correct test depends on your research question: use one-tailed when you have a directional hypothesis and two-tailed when you are testing for any difference.

Work through these three examples to see how z-scores are calculated and interpreted in real scenarios involving exam scores and percentile lookups.