The Normal Distribution AA SL

The normal distribution is the bell-shaped model for continuous measurements such as heights, masses and times, and every question about it is an area under the curve. This page shows how to find a probability from a value, how to find a value from a probability, and how to use z-values to find an unknown mean or standard deviation, with worked examples and exam-style questions.

Which method to use

Imagine the bell curve and the area you want, then choose the calculator function.

The question asks forUseNotes
A probability between two valuesNormal cumulative distribution with the two values, the mean and the standard deviationEnter the standard deviation, not the variance. If the question gives N(50,82)N(50,8^2), type 8, not 64
P(X>a)P(X>a) or P(X<a)P(X<a)Normal cumulative distribution with a very large number (or a very small one) as the missing limitFor example lower limit aa and upper limit 109910^{99}. Because P(X=a)=0P(X=a)=0, the inequalities << and ≤\le give the same answer
A probability within a whole number of standard deviations of the meanThe 68%, 95% and 99.7% ruleAbout 68% lies within μ±σ\mu\pm\sigma, 95% within μ±2σ\mu\pm2\sigma and 99.7% within μ±3σ\mu\pm3\sigma. The rule gives an estimate, so use the calculator when exact values are needed
A probability or value that mirrors a given one, or a mean from two equally likely valuesSymmetry about the meanValues the same distance either side of μ\mu have equal tail areas, and P(X<μ)=0.5P(X<\mu)=0.5. If P(X<a)=P(X>b)P(X<a)=P(X>b) then μ=a+b2\mu=\dfrac{a+b}{2}
The number of items in a group that fall in a regionThe probability of the region multiplied by the group sizeThe result is an estimate, so give a sensible whole number
A value kk for a given probability, with the mean and standard deviation knownThe inverse normal function with the area to the left of kk, the mean and the standard deviationIt uses the area to the left. For an upper tail such as "the top 10%" use 1−0.1=0.91-0.1=0.9
A value kk with a given probability between μ−k\mu-k and μ+k\mu+kEqual tails, then the inverse normal functionEach tail has area 1−p2\dfrac{1-p}{2}, so the upper boundary has area p+1−p2p+\dfrac{1-p}{2} to its left
An unknown mean or an unknown standard deviation, given one probabilityInverse normal on the standard normal (mean 0, standard deviation 1) to find zz, then z=x−μσz=\dfrac{x-\mu}{\sigma}Convert to an area on the left first. Then solve the equation for the unknown
Both the mean and the standard deviation unknownTwo probabilities give two zz-values and two equations in μ\mu and σ\sigmaSolve the pair of equations simultaneously. One probability can only find one unknown
How far a value is from the mean, or a comparison of values from different distributionsThe zz-value z=x−μσz=\dfrac{x-\mu}{\sigma}The zz-value is the number of standard deviations above (positive) or below (negative) the mean. The larger zz is further above its own mean

Section 1 of 2 SL4.9

Normal probabilities

The normal distribution X∼N(μ,σ2)X\sim N(\mu,\sigma^2) is a symmetric, bell-shaped model for continuous data, with mean μ\mu and standard deviation σ\sigma. Approximately 68%, 95% and 99.7% of the values lie within one, two and three standard deviations of the mean. Probabilities are areas under the curve, found using technology.

Notation and the empirical rule

Normal probability and inverse normal calculations are found using technology.

Notation

X∼N(μ,σ2)X\sim N(\mu,\sigma^2)

μ\mu is the mean and σ2\sigma^2 is the variance. This is standard notation rather than a formula, so it isn't listed in the formula booklet.

68-95-99.7 rule

≈68%\approx 68\% within μ±σ\mu\pm\sigma, 95%95\% within μ±2σ\mu\pm2\sigma, 99.7%99.7\% within μ±3σ\mu\pm3\sigma.

Handy for a quick sanity check on a normalcdf answer, and for questions phrased around whole numbers of standard deviations.

Key terms

What do μ\mu and σ\sigma represent?

μ\mu is the mean, the centre of the curve, and σ\sigma is the standard deviation, which sets how wide the curve is. In N(μ,σ2)N(\mu,\sigma^2) the second number is the variance.

e.g. X∼N(50,82)X\sim N(50,8^2) has mean 5050, standard deviation 88 and variance 6464.

What does the area under the curve mean?

The total area under the curve is 11, and the area between two values is the probability that XX lies between them. For a continuous variable, P(X=a)=0P(X=a)=0.

e.g. P(X<μ)=0.5P(X<\mu)=0.5, because half of the area lies to the left of the mean.

How does symmetry help find a probability?

The curve is symmetric about μ\mu, so values at equal distances either side of the mean have equal tail areas. Use this to find a probability from its mirror image.

e.g. If P(X>58)=0.159P(X>58)=0.159 for X∼N(50,82)X\sim N(50,8^2), then P(X<42)=0.159P(X<42)=0.159 too, since 4242 and 5858 are each 88 from the mean.

What happens when the mean or standard deviation changes?

Changing μ\mu slides the curve along the horizontal axis without changing its shape. Increasing σ\sigma makes the curve wider and lower.

e.g. N(60,52)N(60,5^2) has the same shape as N(40,52)N(40,5^2) but is centred at 6060, while N(40,102)N(40,10^2) is wider and lower.

Worked example, step by step

The question Calculator needed [4 marks]

The mass, MM grams, of a packet of rice is normally distributed with mean 505505 and standard deviation 66.

(a) Find the probability that a packet has a mass between 500500 g and 512512 g.

(b) A packet is rejected if its mass is more than 88 g away from the mean. Find the probability that a packet is rejected.

Worked solution: worked example

Answer: (a) 0.6760.676. (b) 0.1820.182.

Step 1 - Write the distribution.

M∼N(505,62)M\sim N(505,\,6^2). The mean is 505505 and the standard deviation is 66.

Why? Every normal question starts from the mean and the standard deviation. The notation gives the variance 626^2, and the calculator needs the standard deviation, 66, so be careful to enter 66 and not 3636.

Step 2 - (a) Set up the probability.

The required probability is P(500<M<512)P(500<M<512), the area under the curve between 500500 and 512512.

Why? A normal probability is an area, so name the two limits of the region first. Here 500500 is below the mean and 512512 is above it, so the region contains the centre of the curve.

Step 3 - (a) Use the calculator.

Normal cumulative distribution with lower limit 500500, upper limit 512512, mean 505505 and standard deviation 66: P(500<M<512)=0.676.P(500<M<512)=0.676. M1 - Use the normal distribution with mean 505505, standard deviation 66 and limits 500500 and 512512 A1 - 0.6760.676

Why? The calculator finds the area between two limits directly, so no standardising is needed. The answer is more than 0.50.5, which is sensible because the region includes the mean and stretches about one standard deviation either side of it.

Step 4 - (b) Describe the rejected region.

Rejected packets have a mass below 505−8=497505-8=497 or above 505+8=513505+8=513. The accepted packets are those between 497497 and 513513, so P(rejected)=1−P(497<M<513).P(\text{rejected})=1-P(497<M<513). M1 - Use 1−P(497<M<513)1-P(497<M<513), or find both tails

Why? "More than 88 g away" means on either side of the mean, so there are two tails. The two limits are the same distance from the mean, which makes the region symmetric. It is quicker to find the accepted region in one calculation and subtract from 11 than to find two tails separately.

Step 5 - (b) Evaluate.

Using the calculator, P(497<M<513)=0.8176…P(497<M<513)=0.8176\ldots, so P(rejected)=1−0.8176…=0.182.P(\text{rejected})=1-0.8176\ldots=0.182. Alternatively, by symmetry, 2×P(M>513)=0.1822\times P(M>513)=0.182. A1 - 0.1820.182

Why? Keep the full calculator value until the last line so the rounding is right. The symmetric tails each have the same area, which gives a second way to reach the same answer.

Step 6 - Check.

The limits 497497 and 513513 are 86≈1.33\dfrac{8}{6}\approx1.33 standard deviations from the mean. About 68%68\% of values lie within 11 standard deviation and about 95%95\% within 22, so the rejected proportion should be between 5%5\% and 32%32\%. The value 0.1820.182 is between them.

Why? The 68%, 95% and 99.7% rule is a fast way to see whether a calculator answer is sensible. It catches mistakes such as typing the variance instead of the standard deviation.

Marks (4):
M1 Use the normal distribution with mean 505505, standard deviation 66 and limits 500500 and 512512
A1 0.6760.676
M1 Use 1−P(497<M<513)1-P(497<M<513), or find both tails
A1 0.1820.182

Now you try: enter your answers to be marked

These are real practice questions. Type your answer in each box and press Mark it. The worked solution and mark scheme appear straight away, and your first attempt is the one that counts. Use the keyboard for powers, roots and trig, and the working box for your notes.

1
Medium
Calculator
[2 marks]
X∼N(50,82)X \sim N(50, 8^2).

Find P(X<60)P(X < 60).
[2]
Fractions & powers
Other

Worked solution: practice question 1

2
Medium
No calc
[6 marks]
The heights of adults in a large group are normally distributed with a mean of 170170 cm and a standard deviation of 1010 cm. A person is chosen at random from the group.
(a) Estimate the probability that the person has a height between 160160 cm and 180180 cm. Give your answer as a decimal correct to 2 decimal places.
[2]
(b) Estimate the probability that the person has a height greater than 180180 cm. Give your answer as a decimal correct to 2 decimal places.
[2]
(c) Estimate the probability that the person has a height between 150150 cm and 190190 cm. Give your answer as a decimal correct to 2 decimal places.
[2]
Fractions & powers
Other

Worked solution: practice question 2

3
Medium
No calc
[5 marks]
The random variable XX is normally distributed with mean μ\mu and standard deviation σ\sigma. It is given that P(X<20)=P(X>36)\mathrm{P}(X<20)=\mathrm{P}(X>36) and that P(X<20)=0.15\mathrm{P}(X<20)=0.15.
(a) Find the value of μ\mu.
[2]
(b) Find P(20<X<36)\mathrm{P}(20<X<36).
[2]
(c) Find P(X>20)\mathrm{P}(X>20).
[1]
Fractions & powers
Other

Worked solution: practice question 3

Your score so far: 0 out of 13 marks

Common mistakes

  • Entering variance instead of standard deviation. The GDC's normalcdf and invNorm functions want σ\sigma, not σ2\sigma^2 - if a question states X∼N(50,82)X\sim N(50,8^2), the value to type in is 8, not 64.
  • Forgetting the curve is symmetric. P(X<μ)=0.5P(X<\mu)=0.5 and areas either side of a symmetric interval like μ±σ\mu\pm\sigma are equal - spotting this shortcut saves a full calculator entry.
  • Using the wrong sign of infinity for one-sided regions. "Less than" needs a very negative lower bound (e.g. −1E99-1\text{E}99); "greater than" needs a very large upper bound (e.g. 1E991\text{E}99) - mixing these up silently gives the complement of the intended answer.

Try some questions on normal probabilities

Section 2 of 2 SL4.12

Inverse normal and unknown parameters

An inverse normal calculation works backwards from a probability to a value of XX. Given P(X<k)=pP(X<k)=p for X∼N(μ,σ2)X\sim N(\mu,\sigma^2), use technology to find kk. When μ\mu or σ\sigma is unknown, first find the standardised value zz from the probability, then solve z=x−μσz=\dfrac{x-\mu}{\sigma} for the missing parameter.

Notation and standardising

Notation

X∼N(μ,σ2)X\sim N(\mu,\sigma^2)

μ\mu is the mean and σ2\sigma^2 is the variance. This is standard notation rather than a formula, so it isn't in the formula booklet.

Standardising

z=x−μσz = \dfrac{x-\mu}{\sigma}

In the formula booklet, and essential when μ\mu or σ\sigma is unknown - rearrange this equation once you know zz from the inverse normal function.

Key terms

What is the left-tail area?

It is the probability P(X<k)P(X<k), the area to the left of the value kk. Inverse normal technology takes this area as its input.

e.g. For P(X>k)=0.2P(X>k)=0.2, the left-tail area is 1−0.2=0.81-0.2=0.8.

What is a zz-value?

A zz-value is the number of standard deviations a value lies from the mean, z=x−μσz=\dfrac{x-\mu}{\sigma}. It is positive above the mean and negative below it.

e.g. For N(70,62)N(70,6^2), the value x=79x=79 has z=79−706=1.5z=\dfrac{79-70}{6}=1.5.

How do you find kk for an interval symmetric about the mean?

If P(μ−k<X<μ+k)=pP(\mu-k<X<\mu+k)=p, the two tails are equal, so each has area 1−p2\dfrac{1-p}{2}. The upper boundary then has left-tail area p+1−p2p+\dfrac{1-p}{2}.

e.g. If P(μ−k<X<μ+k)=0.9P(\mu-k<X<\mu+k)=0.9, each tail has area 0.050.05, so the upper boundary has left-tail area 0.950.95.

How many unknowns can one probability find?

One probability gives one equation, so it can find only one unknown. If both μ\mu and σ\sigma are unknown, you need two probabilities and two equations.

e.g. P(X<30)=0.9P(X<30)=0.9 alone cannot fix both unknowns, but together with P(X<20)=0.3P(X<20)=0.3 it gives two equations in μ\mu and σ\sigma.

Worked example, step by step

The question Calculator needed [5 marks]

The time, TT minutes, taken to cook a ready meal is normally distributed with mean 2424 and standard deviation 2.52.5.

(a) The slowest 5%5\% of meals take longer than tt minutes. Find the value of tt.

(b) A different meal has cooking time W∼N(μ,42)W\sim N(\mu,\,4^2) minutes, and P(W<30)=0.8P(W<30)=0.8. Find the value of μ\mu.

Worked solution: worked example

Answer: (a) t=28.1t=28.1 minutes. (b) μ=26.6\mu=26.6.

Step 1 - (a) Convert to an area on the left.

The slowest 5%5\% lie above tt, so P(T>t)=0.05P(T>t)=0.05. The area to the left of tt is P(T<t)=1−0.05=0.95.P(T<t)=1-0.05=0.95. M1 - Use a left area of 0.950.95

Why? The inverse normal function always takes the area to the left of the value. A question about the top or slowest part is an upper tail, so subtract from 11 first.

Step 2 - (a) Use the inverse normal function.

Inverse normal with area 0.950.95, mean 2424 and standard deviation 2.52.5: t=28.1.t=28.1. A1 - t=28.1t=28.1

Why? The mean and standard deviation are known, so the calculator gives tt directly and no standardising is needed. The value is above the mean, as it must be when the left area is more than 0.50.5.

Step 3 - (b) Find the z-value.

On the standard normal distribution (mean 00, standard deviation 11), P(Z<z)=0.8P(Z<z)=0.8 gives z=0.8416…z=0.8416\ldots M1 - Find the zz-value for a left area of 0.80.8

Why? The mean is unknown, so the calculator cannot find WW directly. First find how many standard deviations 3030 is from the mean. A left area of 0.80.8 is more than 0.50.5, so zz is positive.

Step 4 - (b) Write and solve the equation.

Use z=x−μσz=\dfrac{x-\mu}{\sigma} with x=30x=30 and σ=4\sigma=4: 0.8416…=30−μ4⇒μ=30−4(0.8416…)=26.6.0.8416\ldots=\dfrac{30-\mu}{4}\quad\Rightarrow\quad\mu=30-4(0.8416\ldots)=26.6. M1 - Use z=x−μσz=\dfrac{x-\mu}{\sigma} to form an equation in μ\mu A1 - μ=26.6\mu=26.6

Why? The standardising formula links the value, the mean and the standard deviation. With only one unknown left, rearranging gives μ\mu. Keep the full value of zz in the calculation.

Step 5 - Check.

Find P(W<30)P(W<30) with mean 26.6335…26.6335\ldots and standard deviation 44. The calculator returns 0.80.8, as given. Also, 3030 is above μ=26.6\mu=26.6, which matches a left area of more than 0.50.5.

Why? Putting the answer back into a normal probability returns the probability you started with. This confirms both the value and that it is on the correct side of the mean.

Marks (5):
M1 Use a left area of 0.950.95
A1 t=28.1t=28.1
M1 Find the zz-value for a left area of 0.80.8
M1 Use z=x−μσz=\dfrac{x-\mu}{\sigma} to form an equation in μ\mu
A1 μ=26.6\mu=26.6

Now you try: enter your answers to be marked

1
Medium
Calculator
[2 marks]
A data set has mean 5050 and standard deviation 8.8.

Find the standardised value (z-score) of 66.66.
[2]
Fractions & powers
Other

Worked solution: practice question 1

2
Medium
Calculator
[4 marks]
X∼N(200,σ2)X \sim N(200, \sigma^2) and P(X>215)=0.1.P(X > 215) = 0.1.

Find σ\sigma.
[4]
Fractions & powers
Other

Worked solution: practice question 2

3
Hard
No calc
[5 marks]
The time, XX minutes, taken by a pupil to complete a puzzle is normally distributed with mean μ\mu and standard deviation σ\sigma. It is given that P(X<12)=0.0228P(X<12)=0.0228 and P(X<30)=0.8413P(X<30)=0.8413. The standardized normal variable is ZZ, and P(Z<−2)=0.0228P(Z<-2)=0.0228 and P(Z<1)=0.8413P(Z<1)=0.8413.
(a) Find the value of σ\sigma.
[3]
(b) Find the value of μ\mu.
[2]
Fractions & powers
Other

Worked solution: practice question 3

Your score so far: 0 out of 11 marks

Common mistakes

  • Entering the wrong area into invNorm. The inverse normal function always wants the area to the left of the boundary - if the question gives a right-tail probability (e.g. "top 15%"), convert it to a left area first by subtracting from 1.
  • Trying to run invNorm with two unknown parameters. The GDC's inverse normal function needs numerical μ\mu and σ\sigma - when one of them is the unknown you're solving for, find zz from the standard normal first, then substitute into z=x−μσz=\frac{x-\mu}{\sigma} by hand.
  • Not rounding a cutoff value the right way. When a boundary represents a real quantity like a mark or a measurement, check whether the context needs rounding up or down - "the minimum mark for a distinction" rounds up, not to the nearest whole number.

Try some questions on inverse normal and unknown parameters

Using your GDC

The exact key presses for the TI-84 Plus CE, Casio fx-CG50, Casio fx-CG100 and TI-Nspire CX, with what the screen shows. Pick your calculator.

Normal distribution probabilities

Used in: Normal probabilities, Inverse normal and unknown parameters

How to find a normal probability on a TI-84 Plus CE

normalcdf (between two bounds)

Press 2nd vars (DISTR) and choose 2:normalcdf(. With the Stat Wizards on, a screen asks for each value in turn: type it and press enter each time.

Enter the lower bound 140, the upper bound 165, the mean 150 and the standard deviation 12, then press enter on Paste.

normalcdf(
lower:140
upper:165
μ:150
σ:12

Press enter to run it. If the Stat Wizards are off, type normalcdf(140,165,150,12) with commas between the four values and press enter.

P(140<X<165)=0.692P(140 < X < 165) = 0.692 to 3 significant figures.

normalcdf(140,165,150,12)
0.6920218

For more than 170, there is no upper limit, so use a very large number as the upper bound: type 1 2nd , (EE) 9 9 for 1E99.

P(X>170)=0.0478P(X > 170) = 0.0478 to 3 significant figures.

normalcdf(170,1E99,150,12)
0.0477904

For less than a value, use −1E99 as the lower bound. If you give only the two bounds, the calculator uses mean 0 and standard deviation 1.

The full guide: Normal distribution probabilities

Inverse normal (find the value for a given probability)

Used in: Inverse normal and unknown parameters

How to find a value from a normal probability on a TI-84 Plus CE

invNorm (area to the left)

Press 2nd vars (DISTR) and choose 3:invNorm(. With the Stat Wizards on, a screen asks for each value in turn: type it and press enter each time.

Enter the area 0.9 (the probability to the left of the value), the mean 150 and the standard deviation 12, then press enter on Paste.

invNorm(
area:0.9
μ:150
σ:12

Press enter to run it. If the Stat Wizards are off, type invNorm(0.9,150,12) with commas between the three values and press enter.

P(X<k)=0.9P(X < k) = 0.9 gives k=165.4k = 165.4 to 4 significant figures.

invNorm(0.9,150,12)
165.3786188

For P(X<k)=0.25P(X < k) = 0.25, repeat with area 0.25: invNorm(0.25,150,12).

k=141.9k = 141.9 to 4 significant figures.

invNorm(0.25,150,12)
141.9061230

The area is always the probability to the left of the value. For the top 10%, use 1 − 0.1 = 0.9. If you give only the area, the calculator uses mean 0 and standard deviation 1.

The full guide: Inverse normal (find the value for a given probability)

See the full list of GDC guides for every calculator task.

Syllabus content covered

CodeSyllabus content
SL4.9The normal distribution and curve. Properties of the normal distribution. Diagrammatic representation. Approximately 68% of the data lies between μ±σ\mu\pm\sigma, 95% lies between μ±2σ\mu\pm2\sigma and 99.7% of the data lies between μ±3σ\mu\pm3\sigma. Normal probability calculations: probabilities and values of the variable must be found using technology. Inverse normal calculations: the mean and standard deviation will be given, and this does not involve transformation to the standardized normal variable zz.
SL4.12Standardization of normal variables (zz-values). The standardized value zz gives the number of standard deviations from the mean. Inverse normal calculations where the mean and standard deviation are unknown: use of zz-values to calculate unknown means and standard deviations. Probabilities and values of the variable must be found using technology.

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

Quick questions

What percentage of data lies within one standard deviation of the mean in a normal distribution?

Approximately 68% of the data lies between μ−σ\mu-\sigma and μ+σ\mu+\sigma, 95% within two standard deviations, and 99.7% within three.

What is an inverse normal calculation?

It works backwards from a known probability to find the boundary value kk, using the GDC's inverse normal function with the left-tail area as the input.

How do you find a normal probability on the GDC?

Use normalcdf(lower, upper, mean, standard deviation)\texttt{normalcdf(lower, upper, mean, standard deviation)}. Use a very large or very small bound like 1E991\text{E}99 or −1E99-1\text{E}99 for one-sided regions such as "greater than" or "less than".

How do you find an unknown mean or standard deviation of a normal distribution?

Convert the given probability into a zz-value with the inverse normal function on the standard normal, then substitute into z=x−μσz=\dfrac{x-\mu}{\sigma} and solve algebraically for the missing parameter.

Is the normal distribution discrete or continuous?

It is continuous, so it models measurements such as height or time. Because the probability of any single exact value is 00, P(X<a)P(X<a) and P(X≤a)P(X\le a) are equal.

How do I know whether the boundary value is above or below the mean?

If the left-tail area is greater than 0.50.5, the boundary lies above the mean; if it is less than 0.50.5, it lies below. A left-tail area of exactly 0.50.5 gives the mean itself.

What is the probability of being below the mean?

It is 0.50.5, because the curve is symmetric about the mean, so half of the area lies on each side.

How can I check my inverse normal answer?

Use technology to find the normal probability for your value; it should return the probability you started with. Also check that the value is on the correct side of the mean.

How do I estimate how many items in a group fall in a region?

Multiply the probability of the region by the number of items. If the probability is 0.160.16 and there are 500500 items, expect about 0.16×500=800.16\times500=80 of them.

Do I need the standard normal distribution for every inverse normal question?

No. When μ\mu and σ\sigma are both given, use the inverse normal function with them directly. The standard normal, with μ=0\mu=0 and σ=1\sigma=1, is only needed to find a zz-value when a parameter is unknown.

Ready to practise properly?

Practice questions on every section, marked instantly like the real exam.