ToolSphere

Z-Score Calculator

Computer science

Calculate z-scores, left, right and between normal probabilities, or inverse percentile cutoffs. See a shaded curve and worked statistics examples.

Standardized score

z = 1.5

P(Z ≤ 1.5) = 0.9331928 (93.31928%)

z = (x − mean) / SD = (85 − 70) / 10. A positive z is above the mean.

Probabilities assume a normal population model. Standardizing a value does not make its dataset normal. Displayed answers are rounded numerical approximations.

-4-2024
Horizontal axis: z (standard deviations from the mean). Shading follows the selected area. The drawing shows −4 to 4; calculations include the full tails.

Need a mean or SD first? Use the descriptive statistics calculator. Explore discrete models in the probability distribution calculator.

On this browser only

Give feedback

Did this tool help?

Tell us what worked, what didn’t, or what you’d like next.

Opens Google Forms in a new tab with this tool’s name filled in. No sign-in or email required. Feedback is optional.

How to use Z-Score Calculator

  1. 1Start with the loaded value, mean and standard deviation to see a z-score and left-tail probability.
  2. 2Choose probability mode for left, right or between areas. Enter original values or standard z-scores.
  3. 3Choose inverse percentile and enter a left percentile strictly between 0 and 100 to find a z or original-value cutoff.
  4. 4Read the shaded curve and compare the calculation with your course assumptions. Use Load example to reset.

About Z-Score Calculator

Calculate a z-score and read the normal curve

Use this z-score calculator for introductory statistics homework, a midterm review or a normal-distribution exercise. It starts with a complete example and shows the answer before the controls. You can standardize an original measurement, find the probability to either side of a bound, measure an interval, or work backward from a percentile. The shaded curve follows your selected area so you can check whether the question asks for the left tail, the right tail or the space between two cutoffs.

The calculation runs in your browser. You do not need to upload a dataset or create an account. Enter the mean and standard deviation supplied by your problem, keeping all measurements in the same units. If you only have a list of observations, the descriptive statistics calculator can help you calculate a mean and standard deviation first.

What a z-score means

The formula is z = (x − mean) / SD. Subtracting the mean centers the observation; dividing by the standard deviation expresses its distance in standard-deviation units. Positive scores lie above the mean, negative scores lie below it, and a value equal to the mean has a zero score. A z-score has no measurement unit because the numerator and denominator use the same unit.

Standardization describes relative position even when the data are not normally distributed. Converting that position into a normal probability requires an additional normal-model assumption. A skewed distribution or a dataset with extreme outliers does not become normal just because you calculate z-scores. Check the model stated in your assignment before interpreting an area as an expected proportion.

Worked example using the loaded values

Suppose a statistics exercise models exam scores with mean 70 and standard deviation 10. An observed score of 85 gives z = (85 − 70) / 10 = 1.5. The library calculates a left probability of 0.9331927987311419, displayed as approximately 93.31928%. Under this model, that is the proportion at or below the observed score. It is not the percentage of exam questions the student answered correctly.

The corresponding right probability is 0.06680720126885803, or approximately 6.6807201%. Switch to probability mode, keep original values selected, and choose Right of the bound to see this region shaded. These two areas cover the entire distribution. For continuous models, including or excluding the endpoint does not change the probability because the probability of one exact value is zero.

Left, right and between probabilities

The cumulative distribution function, written Φ(z), gives P(Z ≤ z). A right-tail question asks for P(Z ≥ z). A between question asks for the area from a lower bound to an upper bound. Select the matching region rather than manually subtracting a displayed rounded answer. The library retains its internal precision and evaluates small upper tails directly to reduce cancellation.

For example, enter standard scores with lower bound −1 and upper bound 1. The between probability is 0.6826894921370861, approximately 68.268949%. Reversing the bounds is treated as an input error instead of silently swapping them. Equal bounds give zero area. The drawing shows standard scores from −4 to 4; probabilities still include the full tails beyond the drawing. Its vertical height is density, so read the shaded area rather than treating a curve height as a probability.

Find a percentile cutoff

Inverse mode answers the reverse question: what value has a given proportion below it? Enter a left percentile as a percentage. A percentile of 97.5 corresponds to probability 0.975. The inverse library returns z = 1.959963984540055. With the example mean 70 and standard deviation 10, converting back with x = mean + SD × z gives x = 89.59963984540055, displayed as about 89.59964.

Choose standard scores if you need only z. Choose original values to see both z and x. Percentile endpoints are excluded because the normal distribution extends indefinitely; its extreme endpoints have infinite cutoffs. A right-tail percentage must be converted to its complementary left percentile before using this mode. Read the result sentence to confirm which side the percentile describes.

Precision, tables and course conventions

This calculator evaluates the error-function relationship to the normal CDF numerically and inverts the smaller tail using bisection. It does not select a nearby entry from a rounded classroom table. At z = 1.96 the computed CDF is approximately 0.9750021048517796; at z = 1 it is approximately 0.841344746068543. These agree with the common table readings 0.9750 and 0.8413 after rounding. Display rounding can hide extremely small differences, and sufficiently extreme tails can underflow in browser arithmetic.

Use the standard deviation specified by your exercise. This page does not estimate it from observations, choose a sample correction, calculate a standard error or perform a t-test. For a sample mean, use the sampling distribution and standard error required by your course. Explore binomial and Poisson questions with the probability distribution calculator, or return to the probability and statistics lab for related practice.

Guides and tutorials

Frequently asked questions

How do I calculate a z-score?

Subtract the mean from the observed value, then divide by the standard deviation: z = (x − mean) / SD. Use matching units and a strictly positive SD.

What does a negative z-score mean?

The observation is below the mean. Its absolute z-score gives the distance from the mean in standard deviations.

What is the probability at z = 1.96?

The left probability is approximately 0.9750021, which rounds to 0.9750 in a standard normal table. The right tail is its complement.

What is the probability at z = 1?

The left probability is approximately 0.84134475, which rounds to 0.8413. This is an area under the normal curve, not the curve height.

Can I find probability between two values?

Yes. Select probability mode and Between two bounds. Enter original values with a mean and SD, or enter standardized z-scores directly. The lower bound must not exceed the upper bound.

How do I find a value from a percentile?

Select inverse percentile, enter a left percentile strictly between 0 and 100, and choose original values. The result shows z and x = mean + SD × z.

Is a percentile entered as a decimal or percentage?

Enter a percentage in inverse mode. For example, enter 97.5 for the 97.5th percentile, rather than entering the probability 0.975.

Why does zero standard deviation give no result?

Standardization divides by SD. A zero or negative SD is outside the supported normal model, so the calculator clears the answer until you correct the input.

Does a z-score prove that my data are normal?

No. Standardization changes location and scale, not the distribution shape. Normal probabilities apply only when the normal model is appropriate.

Are the probabilities exact?

They are double-precision numerical approximations evaluated from the normal distribution, rather than rounded table lookups. Displayed results are rounded; extremely small tails can underflow.

Have a suggestion for Z-Score Calculator?Share your feedback ↗