CALTRX

Normal Distribution Calculator

Enter a mean, standard deviation, and value to find the probability, z-score, and percentile rank instantly — covers left-tail, right-tail, and between-values probabilities with full bell curve breakdown.

Enter your values above to see the results.

Tips & Notes

  • ✓Standard deviation must be positive — it cannot be zero or negative. A σ of 0 means all values are identical and there is no distribution to compute.
  • ✓For between-values probability, compute z-scores for both bounds separately: z₁ = (a−μ)/σ and z₂ = (b−μ)/σ, then subtract: P(a ≤ X ≤ b) = Φ(z₂) − Φ(z₁).
  • ✓The 68-95-99.7 rule gives fast mental estimates: about 68% of data falls within 1 standard deviation of the mean, 95% within 2, and 99.7% within 3.
  • ✓A z-score of ±1.96 corresponds to the central 95% of the distribution — which is why 1.96 appears in nearly every 95% confidence interval formula.
  • ✓The normal distribution is a model, not a law. Real data approximates it; always check whether your data is actually bell-shaped before applying normal probabilities.

Common Mistakes

  • ✗Confusing left-tail and right-tail probabilities. P(X ≤ x) and P(X ≥ x) always sum to 1 (since the distribution is continuous). If you need "greater than", compute 1 minus the left-tail result.
  • ✗Using the wrong standard deviation — population σ vs. sample s. For known populations (e.g., IQ tests with defined parameters), use σ. For sample data, use s. The formulas are identical; the interpretation differs.
  • ✗Forgetting to standardize before looking up probabilities. You cannot use a z-table directly with raw x values — always convert x to z = (x−μ)/σ first.
  • ✗Applying normal distribution to data that is not normally distributed. Skewed data, count data, or data with hard boundaries (like 0% to 100%) often violates normality. Always check your data distribution first.
  • ✗Double-counting in between-values probability. P(a ≤ X ≤ b) = Φ(z_b) − Φ(z_a). A common error is adding instead of subtracting: Φ(z_b) + Φ(z_a) gives a number that can exceed 1.

Normal Distribution Calculator Overview

The normal distribution — the bell curve — is the most important distribution in statistics. Heights, exam scores, measurement errors, blood pressure readings, and thousands of other real-world phenomena follow it naturally. Understanding the normal distribution means being able to answer questions like: what percentage of people are taller than 6 feet? What is the probability a product falls within tolerance? This calculator computes probabilities, z-scores, and percentile ranks for any normal distribution.

Standardizing to a z-score:

z = (x − μ) / σ
EX: Heights normally distributed, μ = 68 inches, σ = 4 inches. For x = 74 inches: z = (74 − 68) / 4 = 6/4 = 1.50
Left-tail probability — P(X ≤ x):
P(X ≤ x) = Φ(z) — read from standard normal table or calculator
EX: z = 1.50 → Φ(1.50) = 0.9332 → 93.32% of people are 74 inches or shorter
Right-tail probability — P(X ≥ x):
P(X ≥ x) = 1 − Φ(z)
EX: z = 1.50 → P(X ≥ 74) = 1 − 0.9332 = 0.0668 → 6.68% of people are taller than 74 inches
Between-values probability — P(a ≤ X ≤ b):
P(a ≤ X ≤ b) = Φ(z_b) − Φ(z_a)
EX: P(64 ≤ X ≤ 72) with μ=68, σ=4 → z_a=(64−68)/4=−1.0, z_b=(72−68)/4=1.0 → Φ(1.0)−Φ(−1.0) = 0.8413−0.1587 = 0.6827 (68.27%)
The 68-95-99.7 empirical rule — key reference:
RangeZ-score Range% of DataPractical Meaning
μ ± 1σ−1 to +168.27%Most values fall here
μ ± 2σ−2 to +295.45%Almost all typical values
μ ± 3σ−3 to +399.73%Virtually all values; outliers beyond
μ ± 1.96σ−1.96 to +1.9695.00%Standard 95% confidence interval
The z-score converts any normal distribution into the standard normal (mean 0, standard deviation 1), making probabilities and percentiles directly comparable regardless of the original units or scale. A z-score of +1 means one standard deviation above the mean; z = −2 means two standard deviations below. The empirical rule follows directly: approximately 68% of values fall within 1 SD (z between −1 and +1), 95% within 2 SDs, and 99.7% within 3 SDs. The normal distribution is the natural result of averaging many independent random influences — this is why it appears so consistently across nature, measurement error, and sample statistics. The Central Limit Theorem guarantees that the distribution of sample means approaches normal regardless of the underlying population distribution, as sample size grows. This is why normal-distribution methods remain valid even when individual observations are not normally distributed, provided the sample is large enough.

Frequently Asked Questions

The standard normal distribution has mean μ=0 and standard deviation σ=1. Any normal distribution converts to it using z=(x−μ)/σ. This standardization is why one z-table works for all normal distributions — you convert your specific distribution to the standard one, look up Φ(z), then interpret in your original units. The standard normal is the reference distribution for all z-score calculations and confidence intervals.

Compute z-scores for both values, then subtract their cumulative probabilities: P(a ≤ X ≤ b) = Φ(z_b) − Φ(z_a). Example: IQ scores with μ=100, σ=15. P(85 ≤ X ≤ 115): z₁=(85−100)/15=−1.0, z₂=(115−100)/15=1.0 → Φ(1.0)−Φ(−1.0) = 0.8413−0.1587 = 0.6827. About 68.27% of people score between 85 and 115. Never add the two values — always subtract the lower from the higher.

A z-score of 2.0 means the value is 2 standard deviations above the mean. Φ(2.0)=0.9772, so 97.72% of the distribution falls at or below this point — only 2.28% of values exceed it. For IQ with μ=100, σ=15: z=2.0 corresponds to IQ=130. About 97.72% of people score 130 or below. A z-score of −2.0 means 2 SDs below the mean, with only 2.28% of values below it.

Reverse the z-score process: find z for your percentile using z=Φ⁻¹(percentile), then convert back: x=μ+z×σ. Example: 90th percentile of exam scores with μ=75, σ=10. The z-score for the 90th percentile is 1.2816 → x=75+1.2816×10=75+12.816=87.82. A score of 87.82 is at the 90th percentile — only 10% of students score above it. For the 10th percentile, z=−1.2816 → x=75−12.816=62.18, below which only 10% of students fall.

Data tends to be normally distributed when it results from many small independent random factors adding together — heights, weights, measurement errors, standardized test scores. Data is NOT normally distributed when it is counts (use Poisson), proportions near 0 or 1 (use binomial), time-to-event (use exponential), or heavily skewed. Always visualize with a histogram and check skewness before assuming normality.

They express the same number differently. Left-tail probability P(X≤x)=0.9332 means the same as 'x is at the 93.32nd percentile'. Percentile rank = P(X≤x)×100. A value at the 50th percentile has P(X≤x)=0.50, which equals the mean for any symmetric normal distribution. Both come from the same Φ(z) calculation — percentile just multiplies by 100 to express it as a percentage.