Density, area and notation
Here the second parameter explicitly denotes variance σ2. Some books write N(μ,σ); check the definition being used.[1][2]
The density is nonnegative and has total area one. At every individual x, P(X=x)=0 although f(x)>0. If X is measured in mm, f has unit 1/mm and an integrated area is dimensionless. A density may exceed one, for example for very small σ. This violates no probability rule because probabilities are interval areas.[1]
A symmetric histogram does not prove normality. The generating process, data and required accuracy must support the model choice. Normal distributions extend without bound in both directions; for strictly nonnegative quantities an unsuitable normal model can assign unrealistic negative values. The calculator does not infer normality from data.
Why every normal model uses the same table
The substitution x=μ+σz gives dx=σ dz, cancelling the density factor 1/σ.[1]
For a≤b; open and closed interval endpoints have identical probabilities in this continuous model.[1][2]
The p quantile leaves proportion p to its left. For p=0 or p=1 no finite normal quantile exists.[1]
The exponential formula defines the model; it is not a consequence of μ and σ alone for every distribution. Its integral over the full line is one, following from the Gaussian integral and substitution. Φ has no elementary antiderivative, so tables and calculators use numerical methods. Symmetry gives Φ(0)=1/2.[1]
An interval area and a cutoff
Dimensions in an idealized normal model
- Given
- A length X is assumed normal with μ=50 mm and σ=2 mm.
- Find
- Find the fraction between 48 and 53 mm.
Standardize both endpoints and subtract their cumulative left areas.
Standardize the endpoints separately.
Evaluate left areas at both endpoints.
Subtract to obtain the enclosed area.
Result: Under the normal model about 77.45% lie in this interval.
Check: The excluded areas are about 15.8655% on the left and 6.6807% on the right. Together with the interval area they total 100%. The result plausibly exceeds the central ±1σ area because the right endpoint reaches 1.5σ.
An upper model cutoff
- Given
- A dimensionless quantity has a normal model with μ=100 and σ=15. Find its 90th percentile.
- Find
- Find x0.9 and the remaining right-tail probability.
Convert the specified left area into a standard-normal quantile, then back to the original scale.
Find the standard normal cutoff with left area 0.9.
Transform back to the original scale.
Find the complementary area.
Result: The 90th percentile is about 119.223, leaving 10% to the right.
Check: Standardizing back gives (119.2232735−100)/15≈1.28155157. The cumulative probability there is about 0.9. The cutoff lies above μ, consistent with a left area greater than 1/2.
Recall the method later with the formula’s flashcards.
10 flashcards · English materials
Already have an account? Sign in for flashcardsAvailable from Essential. A ready-made formula set, no AI credits required. Article exercises are not all included.
Five exercises on density and area
Calculate · Foundation
μ=20, σ=4 and x=14. Find z.
Show hint
Subtract the center before dividing by σ.
Learning goal: Find a z score without a table.
Show solution and calculation
z=(14−20)/4=−1.5.
The value is 1.5 standard deviations below μ.
Check: 20+4(−1.5)=14.
Rearrange · Foundation
μ=8 cm, σ=0.5 cm and z=2. What x value corresponds to it?
Show hint
x=μ+σz.
Learning goal: Return to the original scale.
Show solution and calculation
x=8+0.5·2=9 cm.
Check: (9−8)/0.5=2. z has no unit; x and σ both use cm.
Error analysis · Application
A normal model with σ=0.1 has density f(μ)≈3.989 at its mean. Someone calls this an invalid probability.
Show hint
Density is a height with inverse units.
Learning goal: Do not read density as point probability.
Show solution and calculation
f(μ)=1/(0.1√(2π))≈3.989 is a valid density.
Total area is one and P(X=μ)=0.
Only an interval area is a probability between zero and one; density heights need not lie below one.
Interpret · Application
In a normal model what is P(μ−2σ≤X≤μ+2σ)?
Show hint
Standardize to −2 through 2.
Learning goal: Label the 95% rule as an approximation.
Show solution and calculation
Φ(2)−Φ(−2)=2Φ(2)−1≈0.9544997361, about 95.45%. “95%” is a rounded rule of thumb.
Exactly central 95% uses about ±1.959964σ.
Transfer · Challenge
X has μ=30 s and σ=5 s. A cutoff should be exceeded with probability only 0.025. Find it approximately.
Show hint
The left area is 0.975.
Learning goal: Choose a quantile from a right-tail probability.
Show solution and calculation
z=Φ−1(0.975)≈1.959963985.
Thus x=30+5 z≈39.7998199 s.
Check: (x−30)/5≈1.96 and the right area is about 0.025. This is not the 2.5th percentile, which would lie below the mean.
Check your understanding with the formula’s reference test.
8 test questions · English materials
Already have an account? Sign in for the testAvailable from Essential. A ready-made formula set, no AI credits required. Article exercises are not all included.
Numerics and explicit model limits
The calculator standardizes and converts z scores back. It also evaluates density, left area, right tail and quantiles within |z|≤8. Near zero it integrates the exponential power series; the positive tail uses a continued fraction transformed from DLMF 7.9.2. Quantiles use bisection. Results are numerical approximations; the small right tail is evaluated directly to avoid cancellation from one minus almost one.[3]
Expected value and Standard deviation are also defined for many nonnormal models. The Binomial distribution is discrete and gives individual counts positive probability. A Confidence interval for a mean describes an estimation procedure and must not be confused with an interval for individual normally distributed values.
Sources for this explanation
Quanta · Updated November 1, 2026
- NIST Engineering Statistics Handbook §1.3.6.6.1: normal density, CDF and numerical quantiles
- OpenStax Introductory Statistics 2e §6.1: z transformation and empirical normal rule
- NIST DLMF §7.9 equation7.9.2: continued fraction for the complementary error function
Editorial responsibility: Quanta Study editorial team
Review scope
This explanation, 2 worked examples and 5 exercises are available in German and English. Structure, source references and mathematical notation undergo automated checks. Calculations, model assumptions and both language versions receive a separate AI review.
An independent, named human subject review has not yet been documented. Our exercise numbers are worked examples, rather than measurements published by the cited sources.
Editorial responsibility, sources and methodsChange history
- Authored a full German-English article with an original explanation, two worked examples and five varied exercises.
Cite this reference page
Quanta. Normal Distribution (Gaussian Bell Curve). Updated November 1, 2026. https://quanta-study.de/en/formula-sheet/normalverteilung
When citing a specific scientific statement, also cite the specialist source linked to that statement.