STATISTICS CALCULATOR

Z-Score Calculator

Measure how many standard deviations an observation lies from a supplied mean. Normal-distribution areas are shown separately as model-based estimates.

Calculator

YOUR RESULTS
Z-score1.5
Normal area below
93.3193 %
Normal area above
6.6807 %
Normal area beyond ±|z|
13.3614 %

Areas assume a normal distribution; a z-score alone does not establish normality.

Standard normal reference densities
Standard normal reference densities
z-scoreDensity
-41.34e-4
-3.753.53e-4
-3.58.73e-4
-3.252.03e-3
-34.43e-3
-2.759.09e-3
-2.50.02
-2.250.03
-20.05
-1.750.09
-1.50.13
-1.250.18
-10.24
-0.750.3
-0.50.35
-0.250.39
00.4
0.250.39
0.50.35
0.750.3
10.24
1.250.18
1.50.13
1.750.09
20.05
2.250.03
2.50.02
2.759.09e-3
34.43e-3
3.252.03e-3
3.58.73e-4
3.753.53e-4
41.34e-4

Understanding your result

Positive z-scores lie above the mean and negative scores below it. The lower-tail percentage is a modeled percentile, while the two-tail result combines both equally distant extremes. These are not empirical ranks from your dataset.

The formula

z = (observed value − mean) ÷ standard deviation. Normal area below = Φ(z); above = 1 − Φ(z). Two-tail area = 2 × Φ(−|z|).

Subtracting the mean centers the observation and dividing by a positive standard deviation expresses its distance in spread units. The normal CDF is evaluated numerically; displayed percentages round to four decimal places.

Worked example

For an observation of 85, mean 70, and standard deviation 10, z = 1.5. Under a normal model, approximately 93.3193% lies below that value, 6.6807% above it, and 13.3614% beyond either −1.5 or +1.5.

How to use this calculator

  1. Enter the observation and mean using the same measurement scale.
  2. Supply a strictly positive standard deviation in those units.
  3. Read the standardized distance, then interpret areas only if a normal model is appropriate.

Standardization does not prove normality

A z-score is a rescaling operation that can be applied to many datasets. Interpreting it as a normal percentile is an additional modeling decision. Skewed, heavy-tailed, or mixed populations can have very different tail frequencies. The density table describes the assumed normal model; it is not fitted from observations entered here.

Tail areas are not automatically test results

A formal hypothesis test needs its own sampling model and assumptions. For example, uncertainty in an estimated mean or standard deviation may require a different distribution. This tool does not choose a significance threshold or establish causation. Very small normal-tail areas can display as zero after rounding even though their mathematical probability is positive.

Assumptions & limitations

What this calculation assumes

  • Normal areas use a continuous standard normal distribution.

What to keep in mind

  • No normality test, sample-size adjustment, t-distribution, or empirical percentile calculation.

Common questions

Why must standard deviation be positive?

Zero spread makes division undefined. A negative standard deviation is not a valid spread measure.

Does a z-score of zero mean no data?

No. It means the observation equals the supplied mean. The normal reference has half its area on each side.

Sources & further reading