Inverse Normal and Parameters (AA HL)

Inverse normal questions work backwards from a probability to a value. If X∼N(μ,σ2)X\sim N(\mu,\sigma^2) and P(X<k)=pP(X<k)=p is given, then kk is found with the inverse normal function on the GDC, using the area to the left. When μ\mu or σ\sigma is unknown, standardise using z=x−μσz=\dfrac{x-\mu}{\sigma} and solve for it. Upper-tail and symmetric-interval questions need one extra step before the function is used. It's part of the broader Probability Distributions topic.

11 questions on this sub-topic.

Syllabus

CodeSyllabus content
SL4.9The normal distribution and curve, and its properties (approximately 68% of data within 1 standard deviation, 95% within 2, 99.7% within 3). Normal and inverse normal probabilities are found using technology.

Official IB source: Mathematics: analysis and approaches subject brief (PDF, ibo.org).

Skills practised on this page:

The key formula

The standardizing formula below is what links the raw value xx to the standard-normal zz-value your calculator's inverse function actually returns.

Standardizing

z=x−μσz=\dfrac{x-\mu}{\sigma}

Rearrange this to solve for whichever quantity is unknown - xx, μ\mu, or σ\sigma - once you have zz from the GDC.

Inverse normal

Given a probability, find the corresponding xx-value - always found using the GDC's inverse normal function, working from the area to the LEFT of the value.

Need the full syllabus wording and formula-booklet reference table? See Probability Distributions. For calculator steps, see the parent topic's GDC guidance.

Key terms

What is a z-value?

A zz-value is the number of standard deviations a value lies from the mean, z=x−μσz=\dfrac{x-\mu}{\sigma}. The standard normal distribution has mean 00 and standard deviation 11.

e.g. For X∼N(50,102)X\sim N(50,10^2), the value x=65x=65 has z=65−5010=1.5z=\dfrac{65-50}{10}=1.5.

What does X ~ N(mu, sigma squared) tell you?

It says that XX is normally distributed with mean μ\mu and variance σ2\sigma^2, so the standard deviation is σ\sigma.

e.g. X∼N(64,92)X\sim N(64,9^2) has mean 6464 and standard deviation 99, so its variance is 8181.

How do you handle an upper-tail probability?

The inverse normal function works with the area to the left, so subtract an upper-tail probability from 11 first.

e.g. P(X>k)=0.1P(X>k)=0.1 means P(X<k)=0.9P(X<k)=0.9.

How do you find an unknown standard deviation?

Use the inverse normal function on the standard normal distribution to find zz for the given probability, then solve z=x−μσz=\dfrac{x-\mu}{\sigma} for σ\sigma.

e.g. For X∼N(40,σ2)X\sim N(40,\sigma^2) with P(X<46)=0.9P(X<46)=0.9, z≈1.2816z\approx1.2816, so σ=61.2816≈4.68\sigma=\dfrac{6}{1.2816}\approx4.68.

Worked examples

1
Medium
GDC
[4 marks]

The lifetime of a certain battery is modelled by X∼N(500,402)X\sim N(500,40^2) hours.

Find the value of kk such that P(500−k<X<500+k)=0.9.P(500-k<X<500+k)=0.9.

Worked solution

By symmetry each tail has probability 0.05,0.05, so P(X<500+k)=0.95.P(X<500+k)=0.95. M1
z=invNorm(0.95)≈1.6449.z=\text{invNorm}(0.95)\approx1.6449. A1
k=1.6449×40k=1.6449\times40 M1
≈65.8.\approx65.8. A1On the GDC. Find kk such that P(X<k)=0.95P(X<k)=0.95, where X∼N(0, 402)X\sim N(0,\ 40^2). The calculator shows about 65.8.

TI-84 Plus CE: 2nd vars 3:invNorm( area (1+0.9)/2 mean 0 sd 40 enterCasio fx-CG50: MENU Statistics F5 F1 F3 F2 ▼ ▼ Area (1+0.9)/2 EXE σ 40 EXE μ 0 EXE ▼ F1Casio fx-CG100: HOME Distribution Normal OK Tail X ≤ (left tail) any x, then μ 0 OK σ 40 OK Execute OK select p, type (1+0.9)/2 OKTI-Nspire CX: menu Probability Distributions Inverse Normal Area (1+0.9)/2 tab μ 0 tab σ 40 enterFull GDC guide: Inverse normal Topic page

M1 Symmetric tails A1 Correct zz-value M1 k=zσk=z\sigma A1 Correct answer of ≈65.8\approx65.8
2
Hard
GDC
[3 marks]

X∼N(40,σ2)X\sim N(40, \sigma^2) and P(X>50)=0.1.P(X>50)=0.1. Find σ.\sigma.

Worked solution

zz for upper 0.10.1 is ≈1.2816.\approx 1.2816. M1 A1
1.2816=50−40σ⇒σ=101.2816≈7.80.1.2816 = \dfrac{50 - 40}{\sigma}\Rightarrow \sigma = \dfrac{10}{1.2816} \approx 7.80. A1On the GDC. For a normal variable with P(X<50)=0.9 and mean 40, find the standard deviation. The calculator shows about 7.80.

TI-84 Plus CE: 2nd vars 3:invNorm( area 0.9 mean 0 sd 1 enter then use σ = (x − μ) / z with AnsCasio fx-CG50: MENU Statistics F5 F1 F3 F2 ▼ ▼ Area 0.9 EXE σ 1 EXE μ 0 EXE ▼ F1 then use σ = (x − μ) / z with AnsCasio fx-CG100: HOME Distribution Normal OK Tail X ≤ (left tail) any x, then μ 0 OK σ 1 OK Execute OK select p, type 0.9 OK then use σ = (x − μ) / z with AnsTI-Nspire CX: menu Probability Distributions Inverse Normal Area 0.9 tab μ 0 tab σ 1 enter then use σ = (x − μ) / z with AnsFull GDC guide: Find an unknown mean or standard deviation Topic page

M1 InvNorm A1 z≈1.2816z\approx1.2816 A1 Correct answer of ≈7.80\approx7.80
3
Medium
GDC
[6 marks]

Exam scores are modelled by X∼N(64,92).X\sim N(64,9^2). The top 10%10\% of students receive a distinction, and the bottom 20%20\% receive a fail.

(a)  Find the minimum score needed for a distinction.

(b)  Find the maximum score that results in a fail.

Worked solution

(a) Top 10% means P(X<k)=0.9, z=invNorm(0.9)≈1.2816.P(X<k)=0.9,\ z=\text{invNorm}(0.9)\approx1.2816. M1 A1
k=64+1.2816(9)≈75.5,k=64+1.2816(9)\approx75.5, so the minimum score for a distinction is 76.76. A1

(b) z=invNorm(0.2)≈−0.8416.z=\text{invNorm}(0.2)\approx-0.8416. M1 A1
k=64−0.8416(9)≈56.4,k=64-0.8416(9)\approx56.4, so the maximum score for a fail is 56.56. A1On the GDC. (a) Find kk such that P(X<k)=0.9P(X<k)=0.9, where X∼N(64, 92)X\sim N(64,\ 9^2). (b) Find kk such that P(X<k)=0.2P(X<k)=0.2, where X∼N(64, 92)X\sim N(64,\ 9^2).

TI-84 Plus CE: (a) 2nd vars 3:invNorm( area 0.9 mean 64 sd 9 enter Then (b) 2nd vars 3:invNorm( area 0.2 mean 64 sd 9 enterCasio fx-CG50: (a) MENU Statistics F5 F1 F3 F2 ▼ ▼ Area 0.9 EXE σ 9 EXE μ 64 EXE ▼ F1 Then (b) MENU Statistics F5 F1 F3 F2 ▼ ▼ Area 0.2 EXE σ 9 EXE μ 64 EXE ▼ F1Casio fx-CG100: (a) HOME Distribution Normal OK Tail X ≤ (left tail) any x, then μ 64 OK σ 9 OK Execute OK select p, type 0.9 OK Then (b) HOME Distribution Normal OK Tail X ≤ (left tail) any x, then μ 64 OK σ 9 OK Execute OK select p, type 0.2 OKTI-Nspire CX: (a) menu Probability Distributions Inverse Normal Area 0.9 tab μ 64 tab σ 9 enter Then (b) menu Probability Distributions Inverse Normal Area 0.2 tab μ 64 tab σ 9 enterFull GDC guide: Inverse normal Topic page

M1 Convert and find zz A1 Correct zz-value A1 Rounds up to a whole mark M1 InvNorm A1 Correct zz-value A1 Rounds down to a whole mark

Common mistakes

  • Reporting the zz-value instead of the actual xx-value. Inverse normal calculations are asking for a real value in context (a mass, a time, a mark), not the intermediate standardized zz-value used to find it.
  • Entering the wrong tail probability. The GDC's inverse normal function always wants the area to the LEFT - if the question gives you an upper-tail or two-tail probability, convert it first or you'll get the wrong side of the distribution entirely.
  • Rearranging the standardizing equation incorrectly. Once zz is known, solving for μ\mu or σ\sigma means rearranging z=x−μσz=\dfrac{x-\mu}{\sigma} - a sign slip here (especially when zz is negative) is one of the most common ways this topic loses marks.

Using your GDC

Every step below is a real button sequence, not a vague "use your calculator" hint - covering the TI-84 Plus, TI-Nspire, and Casio fx-9860/fx-CG50. Pick your model to filter down to just the steps that apply to you.

Show steps for:
Inverse normal (find the value for a given probability)

The reverse of a normal probability: given a percentage, find the cut-off value (e.g. the mark for the top 10%).

  1. Work out the area to the LEFT of the value you want.
  2. 2nd → VARS (DISTR) → invNorm(area, μ, σ). Newer OS lets you pick the tail.TI-84
  3. menu → Probability → Distributions → Inverse Normal; enter the area, μ and σ.Nspire
  4. Main menu → Statistics → DIST → NORM → InvN; set the tail and enter area, σ, μ.Casio

Tip: invNorm needs the area to the LEFT. For "top 10%", use area = 0.90; for "bottom 25%", use area = 0.25.

Find an unknown mean or standard deviation (normal)

Given a probability and one parameter, work back to the missing μ or σ - a standard exam twist.

  1. Turn the probability into a z-value using the inverse normal with μ = 0, σ = 1.
  2. invNorm(area-to-left, 0, 1) gives z; then solve z=(x−μ)/σz = (x - \mu)/\sigma for the unknown.TI-84
  3. Use Inverse Normal with μ = 0, σ = 1 to get z, then solve the standardising equation for μ or σ.Nspire
  4. DIST → NORM → InvN with μ = 0, σ = 1 gives z; substitute into z=(x−μ)/σz = (x - \mu)/\sigma.Casio
  5. If two probabilities are given, form two equations and solve them simultaneously for μ and σ.

Tip: Sketch the curve and shade the area on the correct side before using invNorm.

See the full GDC guide for more calculator models and topics.

Quick questions

How do you find a value from a normal distribution when you're given a probability?

Use the GDC's inverse normal function with the area to the LEFT of the unknown value. It returns the xx-value directly (or the zz-value if you're working in standard form).

Why does inverse normal always need the area to the left?

That's how the function is defined on every GDC - it treats the number you enter as the cumulative probability up to that point. If you're given an upper-tail or two-tail probability, convert it to a left-tail probability before entering it.

Why is my z-value sometimes negative?

A value below the mean has a negative zz-value. Any left-tail probability below 0.50.5 gives a negative zz.

How do I find two values placed symmetrically about the mean?

For P(μ−k<X<μ+k)=pP(\mu-k<X<\mu+k)=p, the area to the left of μ+k\mu+k is 0.5+p20.5+\dfrac p2. For p=0.8p=0.8, this left area is 0.90.9.

Do I need to standardise before using inverse normal on the GDC?

Not when μ\mu and σ\sigma are known, because you can enter them directly. Standardise only when μ\mu or σ\sigma is unknown.

Ready to practise properly?

11 inverse-normal questions, marked instantly like the real exam.