A z-score is a standardized score, calculated as z = (x − μ)/σ, that tells you how many standard deviations a value falls above (positive z) or below (negative z) the mean, letting you compare values from different distributions and find normal probabilities.
A z-score (also called a standard score) answers one question: how unusual is this value? You compute it as z = (x − μ)/σ, which is just "distance from the mean, measured in standard deviations." A z-score of +2 means the value sits 2 standard deviations above the mean. A z-score of −0.5 means it's half a standard deviation below. Per the CED (1.10.B), a z-score is one example of a standardized score, and standardizing is what lets you compare apples to oranges. A 95 on an easy quiz and an 82 on a brutal one can't be compared directly, but their z-scores can.
The z-score's superpower is that it converts ANY normal distribution into the standard normal distribution (mean 0, standard deviation 1). Once you've standardized, you can use a z-table or your calculator's normalcdf to find the area under the curve, which is the probability or proportion you're after (5.2.A and 5.2.B). That same logic gets recycled in Unit 6, where the test statistic for a proportion is literally a z-score of your sample statistic, telling you how many standard errors p̂ sits from the hypothesized value.
Z-scores are arguably the single most reused tool in AP Statistics. In Unit 1, they support LO 1.10.B (determining proportions and percentiles from a normal distribution) and 1.10.C (comparing relative position within or between data sets). In Unit 5, LO 5.2.A and 5.2.B have you converting values to z-scores to find probabilities, or working backward from an area to a z-score to a raw value. Then Unit 6 turns the z-score into an inference engine. The one-sample z-test for a proportion (6.4.B), the two-sample z-test statistic (6.11.A), and the z* critical value in confidence intervals (6.8.C) are all the same idea wearing different outfits. If you genuinely understand z = (statistic − parameter)/(standard deviation of statistic), you understand the general test statistic formula on the AP formula sheet, and the CED explicitly says you can rebuild every z-procedure formula from it.
Keep studying AP Statistics Unit 9QyrTKQ1AuLyQwom
Normal Distribution (Units 1 & 5)
Z-scores and the normal curve are a package deal. Standardizing converts any normal distribution into the standard normal, so one table (or one calculator function) handles every normal probability problem on the exam. The empirical rule (68-95-99.7) is just z-scores of 1, 2, and 3 in disguise.
Standard Deviation (Unit 1)
The standard deviation is the ruler a z-score uses. Without σ in the denominator, "x − μ" is just a raw distance with units attached. Dividing by σ strips the units away, which is exactly why z-scores from totally different data sets can be compared.
1-Prop Z-Test (Unit 6)
The test statistic in a one-proportion z-test is a z-score of your sample proportion. It measures how many standard errors p̂ falls from the null value p₀. A big |z| means a small p-value, which means your sample would be weird if H₀ were true. That's the entire logic of the test.
Confidence Interval (Unit 6)
The z* critical value in a z-interval is a z-score running in reverse. Instead of asking "how unusual is this value," you pick a confidence level like 95%, find the z-score that traps that central area (1.96), and build your interval as statistic ± z* times the standard error.
Z-scores show up two main ways. First, normal distribution calculations, which appear constantly in multiple choice and in released FRQs like 2017 Q3 (melon diameters from two distributors), 2018 Q6 (systolic blood pressure), and 2022 Q3 (shampoo bottle fill amounts, mean 0.60 L, SD 0.04 L). You'll standardize a value, find an area, or work backward from a percentile to a raw score, like finding the minimum SAT score for the top 5% when μ = 1060 and σ = 195. Show the z-score calculation in your work; naked calculator answers lose communication points. Second, inference. In Unit 6 FRQs you compute a z test statistic, compare the resulting p-value to α, and state a conclusion in context (6.6.A, 6.11.C). Also expect comparison questions where you decide which of two values is "more impressive" by computing both z-scores, straight from LO 1.10.C.
Both measure relative position (LO 1.10.C pairs them for exactly that reason), but they say different things. A z-score tells you distance from the mean in standard deviations; a percentile tells you the percent of data at or below your value. A z-score of 0 is always the 50th percentile in a normal distribution, but in a skewed data set the connection breaks down. Z-scores need a mean and standard deviation to compute; percentiles only need ranks. On the exam, you often translate between them: z-score → area under the normal curve → percentile.
A z-score is calculated as z = (x − μ)/σ and tells you how many standard deviations a value falls above or below the mean.
Positive z-scores mean the value is above the mean, negative z-scores mean it's below, and z = 0 means the value equals the mean.
Z-scores have no units, which is what makes them perfect for comparing values from different distributions, like scores on two different tests.
For normal distributions, converting to a z-score lets you find probabilities as areas under the standard normal curve using a table or calculator.
You can run the formula backward: given a percentile or area, find the z-score, then solve for the raw value x = μ + zσ.
Every z-procedure in Unit 6 (one-prop z-test, two-prop z-test, z-intervals) is built on the same z-score logic: (statistic − parameter) divided by the standard deviation of the statistic.
A z-score measures how many standard deviations a value falls from the mean, calculated as z = (x − μ)/σ. It standardizes values so you can compare across different distributions and find probabilities under the normal curve.
No. A negative z-score just means the value is below the mean, not that anything is wrong. Whether "below the mean" is good or bad depends on context. A negative z-score for golf strokes or race times is actually great.
Standard deviation describes the typical spread of an entire data set; a z-score describes one specific value's position using that spread as the ruler. The standard deviation is the denominator inside the z-score formula.
No. You can compute a z-score for any distribution with a mean and standard deviation, and it still measures relative position. But the z-score-to-probability conversion (using a z-table or normalcdf) only works when the distribution is approximately normal, which is why you check normality conditions in Unit 6.
Yes, in general form. The formula sheet gives the general standardized test statistic (statistic − parameter)/(standard deviation of statistic), plus the standard error formulas. The CED says you don't need to memorize the specific z-test and z-interval formulas because you can rebuild them from those pieces.