Probability Calculator Using Z-Score: How It Works

This calculator finds the probability that a normally distributed value falls below, above, or between given points, using the z-score formula. You enter the population mean, population standard deviation, and one or two raw scores; the calculator returns the z-score and the corresponding probability. The result is the area under the standard normal curve, which is the likelihood of observing a value in that range.

Formula

The z-score formula converts a raw score into a standard score:

z = (x – μ) / σ

• x – the raw score you are evaluating (same unit as the data).
• μ – the population mean (same unit as the data).
• σ – the population standard deviation (same unit as the data).
• z – the number of standard deviations x is from the mean.

Once you have z, the probability is the area under the standard normal curve. That area is found using the cumulative distribution function (CDF), often denoted Φ(z). Most calculators and software compute Φ(z) directly; the formula above uses the error function (erf) as an equivalent way to get the same number.

Worked Example

Using the default values: mean μ = 100, standard deviation σ = 15, lower raw score x� = 85, upper raw score x₂ = 115.

1. Compute z� for x� = 85: z� = (85 – 100) / 15 = –15 / 15 = –1.00.
2. Compute z₂ for x₂ = 115: z₂ = (115 – 100) / 15 = 15 / 15 = 1.00.
3. Find the probability below z� = –1.00. From the standard normal table, Φ(–1.00) ≈ 0.1587.
4. Find the probability below z₂ = 1.00. From the table, Φ(1.00) ≈ 0.8413.
5. Probability between x� and x₂ = Φ(1.00) – Φ(–1.00) = 0.8413 – 0.1587 = 0.6826.

So about 68.3% of values fall between 85 and 115 when the mean is 100 and the standard deviation is 15. This matches the empirical rule for one standard deviation around the mean.

When to Use the Z-Score Probability Calculator

Use this calculator when your data follows a normal distribution (or approximately normal) and you know the population mean and standard deviation. Common situations include test scores, heights, measurement errors, and quality control metrics. It is also appropriate when you have a sample mean and want to find the probability of getting a sample mean that far from the population mean – in that case, use the standard error (σ / √n) instead of σ in the denominator.

Do not use it for skewed distributions, small samples from non-normal populations, or when you only have sample data without knowing the population parameters. For sample data, use the t-distribution instead.

How to Read the Result

The probability output is a decimal between 0 and 1. Multiply by 100 to get a percentage. For example, 0.6826 means 68.26% of the time a randomly selected value will fall in that range. The z-score itself tells you the direction and distance from the mean: a negative z means the raw score is below the mean, a positive z means above.

The calculator shows two probabilities: the probability between the two raw scores (if you entered both) and the probability below the lower raw score. If you only need the probability above a value, subtract the “below� probability from 1. For instance, if the probability below 85 is 0.1587, the probability above 85 is 1 – 0.1587 = 0.8413.

Common Mistakes and How to Avoid Them

• Using sample standard deviation instead of population standard deviation: The z-formula requires the true population σ. If you only have sample data, use the t-distribution.
• Forgetting to use standard error for sample means: When asking “what is the probability of getting a sample mean this far from the population mean?â€�, the denominator must be σ/√n, not σ.
• Mixing up “less thanâ€� and “less than or equal toâ€�: For a continuous distribution, the probability of an exact value is zero, so it does not matter.
• Rounding z too early: If you round z to two decimals and then look up the probability, you may get a slightly different answer. Keep more decimals during calculation.
• Assuming normality without checking: The normal model gives poor probabilities for heavily skewed data. Check a histogram or use a normality test.

Limits of the Z-Score Formula

The z-score formula assumes the population is normally distributed. Real-world data is rarely perfectly normal, but the Central Limit Theorem ensures that sample means become approximately normal for large sample sizes (typically n ≥ 30). For small samples from non-normal populations, the z-approach is unreliable.

Another limit: the formula only works when you know the true population standard deviation. In practice, σ is often unknown and estimated from a sample, which introduces extra uncertainty. The t-distribution accounts for that uncertainty with degrees of freedom.

Finally, the normal distribution extends from negative to positive infinity, so the model can predict probabilities for impossible values (like negative heights) if the mean is close to zero. That is a mathematical artifact, not a practical problem when the mean is several standard deviations from the boundary.

Comparing Z-Score to Other Probability Methods

For normally distributed data, the z-score is the standard method. Alternatives include:

• T-distribution: Use when σ is unknown and estimated from a sample. The t-distribution is wider than the normal, so probabilities are slightly more conservative for small samples.
• Empirical rule: A quick approximation: 68% within 1σ, 95% within 2σ, 99.7% within 3σ. Good for rough estimates, not precise probabilities.
• Nonparametric methods: For non-normal data, use Chebyshev’s inequality (which gives a lower bound) or resampling methods like bootstrapping.

The z-score is the most direct and accurate when its assumptions hold.

Frequently Asked Questions

What is the difference between a z-score and a probability?

A z-score is a standardized value that tells you how many standard deviations a raw score is from the mean. The probability is the area under the normal curve up to that z-score, representing the likelihood of a value being less than or equal to that raw score. You convert one to the other using the standard normal table or a calculator.

Can I use the z-score formula for any data set?

No. The formula assumes the data is normally distributed. If the data is skewed or has heavy tails, the probabilities will be inaccurate. For sample data where the population standard deviation is unknown, use the t-distribution instead.

How do I find the probability that a value is above a certain point?

Calculate the z-score for that point, find the cumulative probability Φ(z) (the area below), and subtract from 1. The result is the probability of a value being greater than that point.

What if I only have sample mean and standard deviation?

If you have a sample mean and sample standard deviation, you can still use the z-formula if the sample size is large (n ≥ 30) and the population is approximately normal. Use the sample standard deviation as an estimate for σ. For small samples, use the t-distribution with n – 1 degrees of freedom.

Why is the probability between two z-scores not simply the difference of the z-scores?

Because the normal curve is not linear. The area under the curve between two z-scores is the difference of the cumulative probabilities, not the difference of the z-scores themselves. For example, between z = –1 and z = 1, the probability is 0.6826, not 2.

Related

• Confidence Interval Calculator Raw Data
• Confidence Interval Difference Calculator
• Chi Square Test Calculator 2×2: How to Use

in mu | Mean | - | 100 | 0
in sigma | Standard deviation | - | 15 | 0
in x1 | Lower raw score | - | 85 | 0
in x2 | Upper raw score (optional) | - | 115 | 0
out z1 | Z-score for x1 | - | (x1 - mu) / sigma | 2
out z2 | Z-score for x2 | - | (x2 - mu) / sigma | 2
out prob_between | Probability between x1 and x2 | - | (0.5 + 0.5 * erf((x2 - mu) / (sigma * sqrt(2)))) - (0.5 + 0.5 * erf((x1 - mu) / (sigma * sqrt(2)))) | 4
out prob_below_x1 | Probability below x1 | - | 0.5 + 0.5 * erf((x1 - mu) / (sigma * sqrt(2))) | 4