4.12 · SLPaper 1Paper 2

Standard normal variables

Builds on 4.9 Normal distribution21 min read

In a minute

A says how many standard deviations a value lies from the mean, so you can compare values from different normal distributions. Every normal variable standardises to the Z∼N(0,1)Z \sim \mathrm{N}(0, 1). That lets you find an unknown mean or standard deviation from a probability, which inverse normal alone can't do.

You'll be able to

  • Standardise a value with z=x−μσz = \frac{x - \mu}{\sigma} and say what its tells you.
  • Use the to find probabilities and numbers of standard deviations.
  • Find an unknown mean or an unknown standard deviation from a given probability, using inverse normal and a zz-value.
  • Find both the mean and the standard deviation from two probabilities by solving two equations.
Before you start2 quick questions on what this builds on
Check yourself
X∼N(30,42)X \sim \mathrm{N}(30, 4^{2}). Use your calculator to find P(X<33)\mathrm{P}(X < 33). Give your answer to 3 s.f.
Check yourself
X∼N(120,225)X \sim \mathrm{N}(120, 225) and P(X>k)=0.1\mathrm{P}(X > k) = 0.1. Find kk.

Revise first: 4.9 Normal distribution

Standardising: the z-value

Ali scores 7474 on a test where the class mean is 6060. Bea scores 7474 on another test where the class mean is 6868. Ali is further above his mean. But who did better compared with their class?

That depends on the spread. Say Ali's class has 1010 and Bea's has 33. Ali is 74−6010=1.4\frac{74 - 60}{10} = 1.4 standard deviations above his mean. Bea is 74−683=2\frac{74 - 68}{3} = 2 above hers. So Bea did better relative to her class.

Counting in standard deviations puts every normal distribution on one scale. That count is the .

In the formula booklet
z=x−μσz = \dfrac{x - \mu}{\sigma}

This is in the formula booklet, so you don't need to memorise it. You do need to know what it means.

The sign tells you the side. z>0z > 0 means above the mean, z<0z < 0 means below it, and z=0z = 0 is the mean itself. The size tells you how far, counted in standard deviations.

Rearranging gives a value from its zz-value: start at the mean and take zz steps of size σ\sigma.

Learn this
x=μ+zσx = \mu + z\sigma
Worked exampleThe masses MM grams of kiwi fruit from a farm are modelled by M∼N(64,82)M \sim \mathrm{N}(64, 8^{2}). Find the zz-value of a kiwi fruit with mass 5252 g.

Try it first, then reveal one step at a time.

  1. Read the model: μ=64\mu = 64 and σ=8\sigma = 8.
Worked exampleChemistry marks CC in a school are modelled by C∼N(66,102)C \sim \mathrm{N}(66, 10^{2}), and Biology marks BB by B∼N(58,62)B \sim \mathrm{N}(58, 6^{2}). Priya scores 8181 in Chemistry. Find the Biology mark bb that is the same number of standard deviations above its mean.

Try it first, then reveal one step at a time.

  1. Standardise Priya's Chemistry mark: z=81−6610=1510=1.5z = \frac{81 - 66}{10} = \frac{15}{10} = 1.5.
Check yourself
The lengths LL cm of a type of leaf are modelled by L∼N(3.2,0.42)L \sim \mathrm{N}(3.2, 0.4^{2}). Find the zz-value of a leaf of length 2.72.7 cm.
Check yourself
Test marks are modelled by X∼N(50,42)X \sim \mathrm{N}(50, 4^{2}). Which statement describes a mark of 4242?

Probabilities with the standard normal

Standardise every value of XX and you get a new variable, ZZ, with the . The mean of XX becomes z=0z = 0, so ZZ has mean 00. A step of one σ\sigma becomes a step of 11, so ZZ has 11.

Try it below with X∼N(140,122)X \sim \mathrm{N}(140, 12^{2}) on top and ZZ underneath. Move xx: the two shaded areas stay equal.

Standardising a normal valueDrag the slider
104116128140152164176−3−2−10123X ~ N(140, 12²)Z ~ N(0, 1)
140
12

161

z = (x − μ)/σ = (161 − 140)/12

z = 1.7500, probability = 0.95994 (P(X < 161.00) = P(Z < 1.7500), x = 161.00)

Subtracting the mean and dividing by σ turns any normal X into the standard normal Z, so P(X < x) = P(Z < z). The two shaded areas are equal.

On Paper 2 you find these areas with technology. Normal cdf gives the same answer with bounds and mean 00, standard deviation 11, as with xx-bounds and μ\mu, σ\sigma.

You need the zz form when a question counts standard deviations from the mean, or when μ\mu or σ\sigma is unknown. On Paper 1, the 68–95–99.7 rule from 4.9 reads P(−1<Z<1)≈0.68\mathrm{P}(-1 < Z < 1) \approx 0.68, and so on.

Worked exampleThe times TT seconds that runners in a club take to run 800800 m are modelled by T∼N(140,122)T \sim \mathrm{N}(140, 12^{2}). Standardise, then use your calculator to find the probability that a runner takes more than 161161 seconds. Give your answer to 3 s.f.

Try it first, then reveal one step at a time.

  1. Standardise the bound: z=161−14012=2112=1.75z = \frac{161 - 140}{12} = \frac{21}{12} = 1.75.
Worked exampleFor a normally distributed variable XX, 90%90\% of values lie within kk standard deviations of the mean. Find kk, to 3 s.f.

Try it first, then reveal one step at a time.

  1. Write the condition with ZZ: P(−k<Z<k)=0.9\mathrm{P}(-k < Z < k) = 0.9. It doesn't matter what μ\mu and σ\sigma are.
Check yourself
Y∼N(80,25)Y \sim \mathrm{N}(80, 25). Which probability equals P(Y<72)\mathrm{P}(Y < 72)?
Check yourself
A variable is normally distributed. Find the probability that a value is more than 2.22.2 standard deviations below the mean. Give your answer to 3 s.f.

Finding an unknown mean or standard deviation

Inverse normal on your GDC needs the mean and the . If one of them is unknown, you can't type it in. But the Z∼N(0,1)Z \sim \mathrm{N}(0, 1) has both known, so find the there first.

1. Sketch the curve. Write the condition as an area to the left, P(X<x)=p\mathrm{P}(X < x) = p.

2. Use with area pp, mean 00 and standard deviation 11. This gives zz.

3. Write the equation x−μσ=z\frac{x - \mu}{\sigma} = z and solve it for the unknown.

4. Check the sign. A value below the mean has a negative zz, and σ\sigma must come out positive.

Learn this
P(X<x)=p  ⟹  x−μσ=z, where P(Z<z)=p\mathrm{P}(X < x) = p \implies \frac{x - \mu}{\sigma} = z, \text{ where } \mathrm{P}(Z < z) = p
Worked exampleThe time BB hours that a phone battery lasts on a full charge is modelled by B∼N(18,σ2)B \sim \mathrm{N}(18, \sigma^{2}). 10%10\% of batteries last less than 1515 hours. Find σ\sigma.

Try it first, then reveal one step at a time.

  1. Write the condition: P(B<15)=0.1\mathrm{P}(B < 15) = 0.1. It is already an area to the left.
Worked exampleA machine fills jars with honey. The mass MM grams of honey in a jar is modelled by M∼N(μ,2.52)M \sim \mathrm{N}(\mu, 2.5^{2}). The machine is set so that 4%4\% of jars hold less than 450450 g. Find μ\mu.

Try it first, then reveal one step at a time.

  1. Write the condition: P(M<450)=0.04\mathrm{P}(M < 450) = 0.04.
Check yourself
X∼N(μ,62)X \sim \mathrm{N}(\mu, 6^{2}) and P(X<40)=0.3\mathrm{P}(X < 40) = 0.3. Find μ\mu. Give your answer to 3 s.f.
Check yourself
Scores SS in a reaction game are modelled by S∼N(72,σ2)S \sim \mathrm{N}(72, \sigma^{2}), and P(S>80)=0.2\mathrm{P}(S > 80) = 0.2. Find σ\sigma.

If both μ\mu and σ\sigma are unknown, you need two facts. Each probability gives one equation in μ\mu and σ\sigma, so solve them together. Subtracting one equation from the other removes μ\mu.

Worked exampleThe lengths LL mm of bolts made by a machine are normally distributed. 6%6\% of bolts are shorter than 48.248.2 mm and 11%11\% are longer than 52.552.5 mm. Use your calculator to find the mean μ\mu and the standard deviation σ\sigma. Give your answers to 3 s.f.

Try it first, then reveal one step at a time.

  1. Lower fact: P(L<48.2)=0.06\mathrm{P}(L < 48.2) = 0.06. Inverse normal with area 0.060.06, mean 00, standard deviation 11 gives z=−1.55477…z = -1.55477\ldots, so 48.2−μσ=−1.55477…\frac{48.2 - \mu}{\sigma} = -1.55477\ldots, which is μ−1.55477…σ=48.2\mu - 1.55477\ldots\sigma = 48.2.

Common mistakes

  • Using the given area on the wrong side in inverse normal. For more than, the area to the left is 1−p1 - p. A wrong side flips the sign of zz, so check it against your sketch.
  • Mixing zz-values and xx-values. A bound given in standard deviations goes with mean 00 and standard deviation 11; a bound in the units of XX goes with μ\mu and σ\sigma.
  • Giving the value xx when the question asks for a number of standard deviations. If it asks for kk standard deviations, the answer is the zz-value.
  • Putting the whole leftover area in one tail. For the middle 90%90\%, each tail holds 5%5\%, so the upper bound has 95%95\% to its left.
  • Rounding zz to 22 or 33 figures before solving for μ\mu or σ\sigma. Keep the full value, or store it, and round only the final answer.

Exam tips

  • Sketch the curve and shade the area for every condition. It tells you the sign of each zz-value and whether the unknown value is above or below the mean.
  • On Paper 2, write down the inverse normal input, such as area 0.10.1 with mean 00 and standard deviation 11, and the equation x−μσ=z\frac{x - \mu}{\sigma} = z before you solve. Those lines earn the method marks.
  • When a later part uses μ\mu or σ\sigma that you found, use the stored values. A wrong earlier value used correctly can still earn follow-through marks.
  • Give answers to 3 s.f. unless the question says otherwise. Check that σ\sigma is positive before you write it down.

Key takeaways

In bookletz=x−μσz = \dfrac{x - \mu}{\sigma}
Learn thisx=μ+zσx = \mu + z\sigma
Learn thisP(X<x)=p  ⟹  x−μσ=z, where P(Z<z)=p\mathrm{P}(X < x) = p \implies \frac{x - \mu}{\sigma} = z, \text{ where } \mathrm{P}(Z < z) = p
Learn thiszz counts standard deviations from the mean: positive above, negative below. Rearranged, x=μ+zσx = \mu + z\sigma.
Learn thisZ∼N(0,1)Z \sim \mathrm{N}(0, 1) is the standard normal, and P(X<x)=P(Z<z)\mathrm{P}(X < x) = \mathrm{P}(Z < z).
Learn thisUnknown μ\mu or σ\sigma: inverse normal with mean 00 and standard deviation 11 gives zz, then solve x−μσ=z\frac{x - \mu}{\sigma} = z.
Learn thisBoth unknown: two probabilities give two equations in μ\mu and σ\sigma. Solve them together.
Learn thisInverse normal wants the area to the left: for P(X>x)=p\mathrm{P}(X > x) = p, use 1−p1 - p.

Practise this

13 past-paper questions on 4.12 in the AA HL bank, each with a mark scheme and a worked explanation.

Practise 4.12 →