Skip to content
thecalculatorstudio

Z-Score Calculator (with percentile and probability)

Enter a value, the mean and the standard deviation. You get the z-score, the percentile, the probability above and below, and the share within and outside that many standard deviations.

Last reviewed

z-score

1.5000

1.5 standard deviations above the mean

-3-2-10123
Below (percentile)
93.3193%
Above
6.6807%
Within ±1.5
86.6386%
Outside both ends
13.3614%

P(Z < 1.5000) = 0.933193 · P(Z > 1.5000) = 0.066807

Probabilities assume the values follow a normal (bell-curve) distribution. A z-score on its own is just a distance in standard deviations, whatever the shape of the data.

What a z-score tells you

A z-score, or standard score, says how far a value is from the mean in units of standard deviation:

z = (x − μ) ÷ σ

where x is the value, μ the mean and σ the standard deviation. A z of 0 is exactly average, +1 is one standard deviation above, and −2 is two below. Because it has no units, z lets you compare things measured on different scales.

Worked example

A test has a mean of 70 and a standard deviation of 10, and you scored 85.

  • z = (85 − 70) ÷ 10 = 1.5.
  • The area under the normal curve to the left of 1.5 is 0.9332, so your score is at about the 93rd percentile: roughly 93.3% of people scored lower.
  • About 6.7% scored higher.

Comparing two tests. You scored 78 in maths (mean 70, standard deviation 8) and 82 in English (mean 74, standard deviation 10). The z-scores are (78 − 70) ÷ 8 = 1.0 and (82 − 74) ÷ 10 = 0.8. Your 78 in maths is the stronger result, even though your English mark is higher.

Finding the value from a z-score

Turn the formula round: x = μ + z × σ. A z of 1.96 on that test is 70 + 1.96 × 10 = 89.6. Choose “Find the value from z” for this.

Finding z from a percentile

To find the cut-off for the top 2.5%, ask for the 97.5th percentile. The answer is z = 1.96. On a scale with mean 100 and standard deviation 15 that is 100 + 1.96 × 15 = 129.4. Choose “Find z from a percentile” for this. The 90th percentile is z = 1.2816.

A short z-table

The table gives the area to the left of z (the percentile as a fraction), the area to the right, and the area in both tails beyond ±z.

z Left of z Right of z Both tails
0 0.5000 0.5000 1.0000
0.25 0.5987 0.4013 0.8026
0.5 0.6915 0.3085 0.6171
0.75 0.7734 0.2266 0.4533
1 0.8413 0.1587 0.3173
1.28 0.8997 0.1003 0.2005
1.5 0.9332 0.0668 0.1336
1.645 0.9500 0.0500 0.1000
1.75 0.9599 0.0401 0.0801
1.96 0.9750 0.0250 0.0500
2 0.9772 0.0228 0.0455
2.33 0.9901 0.0099 0.0198
2.5 0.9938 0.0062 0.0124
2.576 0.9950 0.0050 0.0100
3 0.9987 0.0013 0.0027
3.5 0.9998 0.0002 0.0005

For a negative z, swap “left” and “right”: the curve is symmetric, so the area below −1.5 is the area above +1.5, which is 0.0668.

The 68-95-99.7 rule

In a normal distribution the share of values within 1, 2 and 3 standard deviations of the mean is 68.27%, 95.45% and 99.73%. The “both tails” column above is the rest: at z = 3 only 0.27% of values are further out. Many people treat |z| above 3 as a rule of thumb for an outlier, though the right cut-off depends on the data.

Confidence levels use the same table. A two-sided 90% interval uses z = 1.645, 95% uses 1.96 and 99% uses 2.576, as these leave 10%, 5% and 1% in the two tails. For the mean of a sample, divide the standard deviation by the square root of the sample size first: z = (x̄ − μ) ÷ (σ ÷ √n). A sample mean of 52 from 36 values, with μ = 50 and σ = 6, has a standard error of 6 ÷ 6 = 1, so z = (52 − 50) ÷ 1 = 2.

Population or sample

The formula above uses the population mean μ and standard deviation σ. If you only have a sample, use the sample mean and sample standard deviation instead; the z-score is then an estimate. The statistics calculator and average calculator give both from a list of numbers.

Common mistakes

  • Reading the percentile off a skewed distribution. The table assumes a bell curve. Income, house prices and waiting times are skewed, and their percentiles will not match.
  • Using the variance instead of the standard deviation. Divide by σ, not σ².
  • Mixing up the tails. “Left of z” is the percentile. For a negative z the right-hand area is the large one.
  • Dropping the sign. The sign says above or below the mean, which matters.
  • Rounding z too early. The difference between 1.96 and 2 is more than 2 percentage points of area.

How it is calculated

z is worked in exact decimal arithmetic. The area to the left of z is 1/2 plus the sum of a series for the error function, (1 ÷ √(2π)) × Σ (−1)ⁿ z²ⁿ⁺¹ ÷ (2ⁿ n! (2n + 1)), added until the terms are smaller than 10⁻⁹⁵, so every digit shown is correct. For a percentile, a starting estimate is refined with Newton’s method until it is correct to more than 30 digits. Probabilities are shown for z between −10 and 10; beyond that the tail is under 1 in 10²³.

Frequently asked questions

What is a z-score?

The number of standard deviations a value is from the mean: z = (x − mean) ÷ standard deviation. A score of 85 on a test with mean 70 and standard deviation 10 has z = (85 − 70) ÷ 10 = 1.5, so it is one and a half standard deviations above average.

How do I turn a z-score into a percentile?

Find the area under the normal curve to the left of z. For z = 1.5 that is 0.9332, so the value is at about the 93rd percentile: about 93.3% of values are lower. The calculator gives this, and the area above, directly.

What z-scores go with 90%, 95% and 99% confidence?

For two-sided intervals the critical values are z = 1.645 for 90%, 1.96 for 95% and 2.576 for 99%. For example, 95% of a normal distribution lies within 1.96 standard deviations of the mean.

What does a negative z-score mean?

The value is below the mean. A z of −1.5 is 1.5 standard deviations below average, and its percentile is 6.68%, since the curve is symmetric: the area below −1.5 equals the area above +1.5.

What is the 68-95-99.7 rule?

In a normal distribution about 68.27% of values lie within 1 standard deviation of the mean, 95.45% within 2 and 99.73% within 3. A z-score beyond ±3 is rare and is often treated as an outlier.

Do the percentiles work for any data?

The z-score always works, as it only measures distance in standard deviations. The percentile and probability figures assume the data follow a normal (bell-shaped) distribution. For skewed data they can be well off, so check the shape first.

Related calculators

Sources