4.9 Normal distribution

4.9 Normal distribution

IB Maths: Applications and Interpretation SL: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • X∼N(μ,σ2)X\sim N(\mu,\sigma^2) is a symmetric bell curve with mean, median and mode all equal to μ\mu.
  • The total area under the curve is 1, and P(X=a)=0P(X=a)=0 for any single value.
  • About 68%, 95% and 99.7% of values lie within 11, 22 and 33 standard deviations of the mean.
  • Normal cdf finds a probability from a value; inverse normal finds a value from the area to its left.
  • A larger σ\sigma gives a wider, flatter curve.

The normal curve

A normal distribution is a symmetric bell curve centred on the mean, with total area 1.

A normal distribution models a continuous variable whose values cluster symmetrically around a mean. We write X∼N(μ,σ2)X\sim N(\mu,\sigma^2), where μ\mu is the mean and σ\sigma is the standard deviation.

The curve is a bell shape, symmetrical about x=μx=\mu, so mean, median and mode are all μ\mu. The total area is 11, and the curve never touches the axis. A larger σ\sigma gives a wider, flatter curve.

Because the variable is continuous, P(X=a)=0P(X=a)=0, so P(X<a)=P(X≤a)P(X<a)=P(X\le a).

−6−4−22460.10.20.30.40.50.60.70.80.9xyNormal curve
The normal curve N(m, s²): change the mean and the standard deviation.

A normal curve has its mean at 50. Where is its median?

The 68-95-99.7 rule

About 68%, 95% and 99.7% of values lie within one, two and three standard deviations of the mean.

For any normal distribution the proportions within a given number of standard deviations are fixed. By symmetry, the area splits equally between the two tails.

Masses are N(500,82)N(500,8^2). Between 484484 and 516516 is μ±2σ\mu\pm2\sigma, so about 95%95\% lie there. Above 508=μ+σ508=\mu+\sigma lies half of the remaining 32%32\%, which is 16%16\%.

The rule gives quick estimates; for exact values use technology.

−4−3−2−112340.10.20.30.40.5xymeanN(0, 1)
About 68% of the area lies within one standard deviation of the mean (shaded).
  • μ±σ\mu\pm\sigmaabout 68%
  • μ±2σ\mu\pm2\sigmaabout 95%
  • μ±3σ\mu\pm3\sigmaabout 99.7%

Masses are N(500,82)N(500,8^2). About what percentage lie between 484 and 516?

Normal probability calculations

Use normal cdf with a lower bound, an upper bound, the mean and the standard deviation.

Find probabilities for X∼N(μ,σ2)X\sim N(\mu,\sigma^2) with your GDC (normal cdf): lower bound, upper bound, μ\mu, σ\sigma.

  • P(a<X<b)P(a<X<b): bounds aa and bb
  • P(X<b)P(X<b): lower bound a very small number such as −1099-10^{99}
  • P(X>a)P(X>a): upper bound a very large number such as 109910^{99}

For T∼N(42,62)T\sim N(42,6^2), P(T>50)=0.0912P(T>50)=0.0912 and P(40<T<45)=0.322P(40<T<45)=0.322. Give 3 s.f. and sketch the shaded area to check it is sensible.

2025303540455055600.010.020.030.040.050.060.070.08xyN(42, 36)
T ~ N(42, 6²): the shaded area is P(T > 50) = 0.0912.

Worked example

T∼N(42,62)T\sim N(42,6^2). Find P(40<T<45)P(40<T<45).

X∼N(10,4)X\sim N(10,4). P(X>10)P(X>10) is...

Inverse normal calculations

Use inverse normal with the area to the left of the value, together with the mean and standard deviation.

An inverse normal calculation works backwards: you know a probability and want the value. You do not need to convert to zz.

The GDC uses the area to the left. If P(X<k)=pP(X<k)=p, enter pp; if P(X>k)=pP(X>k)=p, enter 1−p1-p.

For T∼N(42,62)T\sim N(42,6^2), the longest 10%10\% of journeys take more than kk minutes: P(T<k)=0.9P(T<k)=0.9, so k=49.7k=49.7. For M∼N(62,32)M\sim N(62,3^2), the lightest 5%5\% of eggs are under 57.157.1 g.

Check the side: a top percentage gives a value above μ\mu.

2025303540455055600.010.020.030.040.050.060.070.08xyk = 49.7N(42, 36)
Inverse normal: the shaded area to the left of k is 0.9, so k = 49.7.

The top 5% of scores lie above kk. What area do you enter?

The normal model in context

Define the variable, write the model, sketch and shade, use the GDC, then state the result in words.

Modelling questions ask you to interpret results. Use the same plan: define the variable, write X∼N(μ,σ2)X\sim N(\mu,\sigma^2), sketch and shade, then use your GDC.

  • With the same mean, the larger σ\sigma gives more values in both tails.
  • About 95%95\% of values lie within μ±2σ\mu\pm2\sigma.
  • Proportion above or below a limit: normal cdf. Limit for a proportion: inverse normal.

Battery lifetimes: A is N(120,152)N(120,15^2), B is N(120,252)N(120,25^2). The probability above 150150 hours is 0.02280.0228 for A and 0.1150.115 for B. A normal model is an approximation, so state results as estimates.

4060801001201401601802000.0050.010.0150.020.0250.03xyA: N(120, 15²)B: N(120, 25²)
Same mean, different spread: B (dashed) has the larger standard deviation.

Two normal models have the same mean. Which has more very high and very low values?

Try an exam question

Heights of adult men are modelled by H∼N(170,82)H\sim N(170,8^2) cm. Find the probability that a man is taller than 185185 cm. The tallest 10%10\% are taller than kk cm. Find kk.

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Use the 6868–9595–99.799.7 rule to find the percentage of packets with a mass between 484484 g and 516516 g.

1 mark, marked against the mark scheme

Revision notes