The normal distribution — the familiar bell-shaped curve — is the most widely used probability model in statistics. Heights, test scores, measurement errors, monthly sales, and many biological and economic quantities cluster around a mean and taper off symmetrically on both sides. This calculator lets you plug in a mean, a standard deviation, and one or two boundary values, and instantly returns the z-score, the percentile rank, the area below and above your point, the area between two bounds, and the probability density at that point.
Whether you are grading an exam, benchmarking a process, computing a p-value, or teaching the empirical rule, this tool turns abstract integrals into percentages you can actually use.
How to Use
- Enter the Value (x) — the boundary on the bell curve you want to evaluate. On a test this is a raw score; on a production line it might be a measured diameter.
- Optionally set the Upper Bound (x₂). When you provide a second value, the calculator also reports the probability of landing between x and x₂. Leave it blank if you only care about a single tail.
- Enter the Mean (μ) — the center of the distribution. On a test this is the class average; in business it might be the historical monthly average.
- Enter the Std Deviation (σ) — how spread out the values are. It must be greater than zero; a zero standard deviation means every observation equals the mean and the curve collapses to a spike.
- Read the result card. The Z-Score tells you how many standard deviations x sits from the mean. The Percentile Rank is the share of the distribution below x. The Area Above x is the upper tail. The Area Between x and x₂ is the slice between your two bounds. The Probability Density at x is the height of the bell curve at that point.
A positive z-score means x is above the mean; a negative z-score means it is below. Roughly 68% of observations fall within ±1σ of the mean, 95% within ±2σ, and 99.7% within ±3σ — this is the empirical rule, and it is printed on every result.
The Normal Distribution Formulas
Given a mean μ and standard deviation σ, any value x is standardized to a z-score:
z = (x − μ) / σ
The cumulative distribution function (CDF) Φ(z) gives the probability that a normal variable is at or below x:
Φ(z) = 0.5 × (1 + erf(z / √2))
The probability density function (PDF) — the height of the bell curve — is:
f(x) = exp(−z² / 2) / (σ × √(2π))
The area between two bounds x and x₂ is simply the difference:
P(x ≤ X ≤ x₂) = Φ(z₂) − Φ(z)
These are the exact formulas implemented in SciPy's scipy.stats.norm, and they are the same tables you would look up in a statistics textbook.
The Empirical Rule (68–95–99.7)
For a normal distribution, almost all data lives within three standard deviations of the mean:
- ≈ 68.27% of observations fall within μ ± 1σ.
- ≈ 95.45% of observations fall within μ ± 2σ.
- ≈ 99.73% of observations fall within μ ± 3σ.
That is why values beyond μ ± 3σ are called "three-sigma events" — they happen less than 0.3% of the time under a normal model. The result card reminds you of these bands so you can quickly judge whether your value is typical, unusual, or rare.
Examples
Example 1 — A strong exam result. A test has a mean of 72 and a standard deviation of 8. A student scores 85, and we want to know what share of the class scored between 85 and 90. z = (85 − 72) / 8 = 1.625. The percentile rank is 94.79%, meaning about 94.79% of students scored at or below 85, and 5.21% scored above. The area between 85 and 90 is roughly 3.99% — a narrow sliver near the top of the class.
Example 2 — Standard normal (μ = 0, σ = 1). On the unit normal curve, what share of data lies between z = 1 and z = 2? Φ(2) − Φ(1) = 0.9772 − 0.8413 = 13.59%. The percentile at z = 1 is 84.13%, and the upper tail above z = 1 is 15.87%. This is the textbook reference case.
Example 3 — Large-scale process. A factory's monthly output averages 500,000 units with a standard deviation of 250,000 units. One month hits 1,000,000 units, and we ask about the band up to 1,250,000. z = (1,000,000 − 500,000) / 250,000 = 2.0. The percentile rank is 97.72%, only 2.28% of months exceed that level, and the area between 1,000,000 and 1,250,000 is about 2.14%. The engine handles magnitudes beyond 1,000,000 without overflow.
Example 4 — Below the mean. Another student scores 60 on the same test (μ = 72, σ = 8). z = (60 − 72) / 8 = −1.5. The percentile rank drops to 6.68%, and 93.32% of the class scored higher. The sign of z flips the tail, but the absolute magnitude still drives how unusual the result is.
What the Outputs Mean
- Z-Score — signed distance from the mean, in units of σ. z = 0 is exactly average.
- Percentile Rank (Area Below x) — the cumulative probability P(X ≤ x). A percentile of 90% means x is larger than roughly 90% of the distribution.
- Area Above x — the upper tail P(X > x) = 1 − Φ(z).
- Area Between x and x₂ — the probability of landing in the slice between your two bounds.
- Probability Density at x — the height of the bell curve, not a probability by itself; it is highest at the mean and tapers toward zero in the tails.
Remember that all of these figures assume your data are approximately normal. Skewed, bimodal, or heavy-tailed data need different tools.
FAQ
What is a normal distribution in simple terms?
A normal distribution is a symmetric, bell-shaped probability curve centered at its mean μ, with spread controlled by its standard deviation σ. Most observations cluster near the mean and fewer fall far away. It describes heights, measurement errors, standardized test scores, and many natural phenomena.
How do I calculate the probability of x being below a value?
Standardize x to a z-score with z = (x − μ) / σ, then look up the cumulative normal CDF Φ(z). For example, with μ = 72 and σ = 8, a score of 85 gives z = 1.625 and Φ(1.625) ≈ 0.9479, so about 94.79% of observations fall below 85.
What is the empirical rule?
The empirical rule states that for a normal distribution about 68.27% of data lies within one standard deviation of the mean, 95.45% within two, and 99.73% within three. It is a quick sanity check for whether an observation is typical, unusual, or a rare tail event.
What is the difference between the CDF and the PDF?
The CDF (cumulative distribution function) gives the probability that a variable is at or below a given value — always between 0 and 1. The PDF (probability density function) is the height of the bell curve at a point; it can exceed 1 and is used to draw the curve and to compute intervals by integration.
How do I find the probability between two values?
Plug the two bounds into the calculator as x and x₂. The area between is Φ(z₂) − Φ(z), where each z is standardized against the same μ and σ. For the standard normal curve, the area between z = 1 and z = 2 is about 13.59%.
Can z be negative?
Yes. A negative z-score simply means x is below the mean. The sign tells you direction; the absolute value tells you how far into the tail you are. A z of −1.5 corresponds to about the 6.68th percentile.
What if my standard deviation is zero or negative?
A standard deviation cannot be negative. If it is zero, every observation equals the mean, the denominator is zero, and the normal CDF/PDF is undefined. The calculator rejects a non-positive σ rather than returning a meaningless number.
Disclaimer
This normal distribution calculator is provided for informational purposes only and does not constitute statistical, medical, financial, or other professional advice. All probabilities, percentiles, and density values assume an approximately normal distribution and are estimates based on the inputs you enter. Always consult a qualified statistician or relevant professional before making important decisions based on these results.