On this concept: Formula

Not sure which lesson to do first? Find your level and get a plan →

Math · Statistics & Probability · Grade 9-12

Z-Score

A z-score tells how many standard deviations a value is above (+) or below (−) the mean: z=x−μσ.

z=x−μσ

Quick Answer

A z-score tells how many standard deviations a value is above (+) or below (−) the mean: z=x−μσ. Use it to compare values measured on different scales, such as scores on two differently graded tests, or to judge how unusual one value is within its own group. Subtract the mean, then divide by the standard deviation; the sign says whether the value is above or below. For roughly normal data, a standard normal table turns a z-score into a percentile.

Why This Matters

Raw numbers from different scales cannot be compared directly: is 85 on a test with mean 70 better than 32 on a test with mean 28? Measured in standard deviations, with SDs of 10 and 4, they are z=1.5 against z=1, so the 85 is stronger. Z-scores also connect every normal distribution to one standard table, so a single lookup gives percentiles for heights, scores or part sizes.

Intuitive Explanation

A test has mean 70 and SD 10. Ana scores 85, which is 15 points above the mean, and 15 points is one and a half standard deviations: z=85−7010=1.5. The z-score rewrites her score as a distance from the mean, counted in SDs. A score of 60 would be z=−1, one SD below the mean, and a score of exactly 70 would be z=0.

Using the SD as the unit is what makes different tests comparable. Being 15 points above average is remarkable on a test where most scores sit within 5 points of the mean (z=3), and ordinary on one where scores routinely swing by 30 (z=0.5). The z-score builds that context into the number. It also lets every normal distribution share one table: whatever the original mean and SD, z=1.5 falls at about the 93rd percentile.

The raw deviation x−μ looks similar but is not a z-score. It is still in the original units (15 points) and cannot be compared across tests. A percentile is different again: it is the share of values below a point, and for normal data you get it from z, but z=1.5 and the 93rd percentile are different numbers describing the same score. Always standardize with the mean and SD of the value's own group, never with numbers borrowed from another data set.

Think of it like this. Converting to z-scores is like converting prices into one currency before comparing them: dividing by the SD turns 'points on this test' into a shared unit, standard deviations. The conversion is only as good as the data's shape allows; for strongly skewed data, the same z can sit at quite different percentiles.

Core idea. A z-score restates a value as the number of standard deviations it sits above or below the average.


When to Use

Use a z-score when a problem asks how many standard deviations a value is from the mean, how unusual a value is, or which of two values from different scales is more impressive. It is also the first step for a normal-distribution probability whose cut-off is not a whole number of SDs. If the question wants the share of values below a point, convert the z-score to a percentile. If it only asks how far in the original units, the raw difference x−μ is enough.

Tip. Ask: Am I expressing this value as a number of standard deviations from its mean?

How to Recognize It

Before computing a z-score, check that the question is about one value's position within its own group and that you know that group's mean and standard deviation.

  1. Do the values you are comparing come from tests or groups with different averages or spreads?

    If yes, convert each one to a z-score using its own group's mean and standard deviation, then compare the z-scores. If both come from the same group, the raw values already compare fairly.

  2. Does the answer need to be the percent of the group below the value, rather than a distance?

    If yes, the final answer is a percentile; the z-score is the stepping stone, turned into a percent with a normal table when the data is roughly bell-shaped. If a distance is enough, stop at z.

  3. Is the distance you need measured in the data's own units, such as points or centimeters?

    If yes, the raw deviation x−μ answers it and there is no need to divide. Dividing by σ is the step that turns the raw gap into a number of standard deviations.

  4. Is the value itself unknown, with its z-score given?

    If yes, run the formula backwards: x=μ+zσ. A score with z=−1.5 on a test with mean 70 and standard deviation 8 is 70−12=58.

Z-Score compared with Percentile, Standard deviation, and Raw deviation

Z-scores, percentiles, standard deviations and raw deviations all relate a value to its group. Ask what unit the answer is in: standard deviations (z-score), percent of the group (percentile), or data units for the group's spread (standard deviation) or one value's gap (raw deviation).

Z-Score

Meaning
Use this when you want to compare a value's standing across different distributions or look up a normal probability.
Key test
Am I expressing this value as a number of standard deviations from its mean?
Formula
z=x−μσ
Example
Ana scored 85 on a test with mean 70, SD 10. What is her z-score?

Percentile

Meaning
The percent of values at or below a point, read off the cumulative curve.
Key test
Use when you want rank position rather than SD distance.
Example
Scoring in the 90th percentile

Standard deviation

Meaning
The spread itself, in original units — the divisor in the z-formula.
Key test
Use when you want typical spread, not a single value's standing.
Formula
σ=∑(x−μ)2n
Example
Spread of test scores

Raw deviation

Meaning
The unscaled distance x−μ, still in original units.
Key test
Use only as the numerator before dividing by σ.
Formula
x−μ
Example
Score is 15 points above the mean

Formula & Notation

z=x−μσ
z=x−μσ; equivalently, if X∼N(μ,σ2) then Z=X−μσ∼N(0,1)

How to read it: z is the standard score; Z∼N(0,1) is the standard normal distribution

Worked Examples

Example 1 — Z-score of one test score

Easy

Ana scored 85 on a test with mean 70, SD 10. What is her z-score?
Show the steps and the answer
  1. Subtract the mean: 85−70=15 points.

    This raw deviation says how far above average Ana is, but in test points.

  2. Divide by the SD: 1510=1.5.

    Dividing by the spread converts points into standard deviations, a unit that means the same on any test.

  3. Keep the sign: z=+1.5, above the mean.

    A negative z-score would mean below the mean, so the sign is part of the answer.

  4. Check by working back: 70+1.5×10=85.

    Mean plus z times SD must return the original score.

Answer: z=1.5

Takeaway: Subtract the mean, divide by the standard deviation and keep the sign; mean plus z times SD gives back the value.

Example 2 — They want rank, not SDs

Standard

A parent asks 'what percent of students did Ana beat?' rather than how unusual her score is.
Show the steps and the answer
  1. Notice the parent wants a share of students, which is a percentile, not a number of SDs.

    A z-score of 1.5 is a distance; the parent is asking what fraction of scores fall below Ana's.

  2. Treat the scores as roughly normal and look up z=1.5 in a standard normal table: about 0.933.

    The table gives the area under the standard normal curve to the left of z, which is the share of values below.

  3. Convert: 0.933 is about the 93rd percentile.

    Ana scored higher than about 93% of the students.

  4. Estimate with the 68-95-99.7 rule: z=1 is about the 84th percentile and z=2 about the 97.5th, so 93 fits between them.

    A rough estimate like this catches a misread table.

Answer: z 1.5 maps to about the 93rd percentile

Takeaway: A z-score is a distance in SDs; for normal data, a table turns it into a percentile, the share of values below.

Example 3 — Find the mistake: raw points across tests

Application

Mia scored 82 on a math test (mean 75, SD 4) and 88 on an English test (mean 80, SD 10). Mia writes: "I was 7 points above average in math and 8 above in English, so English was my better result." Find the mistake and correct it.
Show the steps and the answer
  1. Name what Mia compared: raw deviations, 82−75=7 and 88−80=8.

    Raw deviations are in each test's own points, and the two tests spread their scores very differently.

  2. Divide each by its own test's SD: zmath=74=1.75 and zEng=810=0.8.

    A value has to be standardized with its own group's mean and SD.

  3. Compare: 1.75>0.8.

    On the math test 7 points is almost two SDs; on the English test 8 points is less than one.

Answer: Math was the stronger result: z=1.75 against z=0.8.

Takeaway: Compare results from different tests with z-scores, not raw differences; a few points matter more when scores are tightly packed.

Where You'll See It

  • Growth charts at a child's checkup

    Doctors compare a child's height with children of the same age and sex. A 10-year-old who is 150 cm tall, in a group with mean 138 cm and SD 6 cm, has z=150−1386=2, taller than about 97.5% of peers. Growth charts are drawn from exactly these comparisons.

  • Flagging a faulty sensor reading

    A factory sensor normally reads 20.0 °C with SD 0.5 °C. A reading of 22.1 °C has z=22.1−20.00.5=4.2. Values more than 3 SDs from the mean are so rare in normal data that monitoring systems flag them automatically for a person to check.

Common Mistakes

  1. The slip: Forgetting to divide by σ

    Instead: x−μ alone is a raw deviation, not a z-score.

  2. The slip: Dropping the sign

    Instead: a negative z means below the mean; the sign carries direction.

  3. The slip: Using the wrong distribution's mean and SD

    Instead: standardize each value with its own data set's μ and σ.


Mini Practice

Try these on your own. Tap Reveal when you want to check.

  1. A student is 180 cm tall. Heights in the class have mean 170 cm and SD 8 cm. What is the z-score?

    Hint: Subtract the mean, then divide by the standard deviation.

  2. A test has mean 74 and SD 6. Find the z-score of a score of 62.

    Hint: Keep the sign: the score is below the mean.

  3. On a test with mean 80 and SD 4, Rosa's z-score is −1.5. What was her raw score?

    Hint: Rearrange z=x−μσ to solve for x.

  4. Maya scored 88 in math (mean 80, SD 5) and 90 in English (mean 84, SD 8). In which subject did she do better compared with her classmates?

    Hint: Find a z-score for each subject and compare.

  5. You want to decide which is more unusual: a 100 m time in seconds or a long jump in metres. Should you compare raw differences from the mean or z-scores?

    Hint: Can you fairly compare seconds with metres directly?

  6. For x=52, μ=60, σ=4, Tom writes: "z=52−60=−8." Correct his work.

    Hint: A z-score counts standard deviations, not raw points.

Want the full set?

50 practice questions for this concept — free to try, every one with a complete worked solution showing the why, not just the answer.

Frequently Asked Questions

How large does a z-score have to be before a value counts as unusual?

A common rule of thumb is beyond 2 in either direction, since in bell-shaped data only about 5% of values are that far out, and beyond 3 for a value worth double-checking as a possible error. The cut-off is a choice, not a law: a factory checking parts may flag anything beyond 2.5, while particle physicists wait for 5 standard deviations before they announce a discovery.

Do doctors use z-scores?

Yes, children's growth is often tracked this way. The World Health Organization's growth standards give, for each age and sex, a typical height and a spread, and a child's height becomes a z-score against them. A height-for-age z-score below −2-2 is the WHO's definition of stunting, a sign of long-term poor nutrition. Using z-scores lets one cut-off work for a 2-year-old and a 5-year-old, even though their typical heights differ.

Can every value in a data set lie within 1 standard deviation of the mean?

Not when you use the population standard deviation. The squared z-scores of a data set always average exactly 1, so at least one value must have ∣z∣≥1|z| \ge 1. For 1, 2 and 3, the mean is 2 and σ=23≈0.82\sigma = \sqrt{\tfrac{2}{3}} \approx 0.82, so the end values have z≈±1.22z \approx \pm 1.22. The standard deviation is a typical distance, and some values have to sit at or beyond it.

Learning Path

Z-Score

You are here

You need the Mean, the reference point, and Standard Deviation, the unit of distance. The z-score adds a scale-free way to place any value relative to its group. Next, the Normal Distribution uses z-scores with the standard normal table to turn a value into a percentile or a probability, such as the share of people taller than a given height.

See Also