Scribe logoScribeNotes & practice
Sign inStart free
  1. Notes
  2. ›IB Mathematics: Applications and Interpretation
  3. ›4 Statistics and probability
  4. ›4.15 Central limit theorem

IB Mathematics: Applications and Interpretation · Topic 4 · Statistics and probability

4.15 Central limit theorem

State the central limit theorem and use it to find probabilities about sample means.

  • 4 min read
  • 6 parts
  • 10 questions
  • Draft
  • HL only
Start a Quick 5 on this topic→Jump to the 10 questions
1/6Sample means vary from one sample to the next
  1. 1Sample means vary from one sample to the next
  2. 2The sample mean keeps the population mean and shrinks the spread
  3. 3For large samples the means follow a normal distribution
  4. 4The theorem turns questions about means into normal probabilities
  5. 5Larger samples give narrower sampling distributions
  6. 6The sample size needed depends on the population shape
  7. Check your recall
  8. Practice: 10 questions
1

Part 1 of 6: Sample means vary from one sample to the next

A teacher records the sleep time of 30 randomly chosen students and finds the mean. Another random group of 30 gives a different mean. The sample mean changes with each sample.

Each sample mean is one value of a random variable, written Xˉ\bar XXˉ. Its probability distribution is the sampling distribution of the sample mean, found from all possible samples of the same size.

The population has its own mean μ\muμ and standard deviation σ\sigmaσ. The sampling distribution links Xˉ\bar XXˉ back to these two values.

Key term

sampling distribution

A sampling distribution is the probability distribution of a statistic, such as the sample mean, over all possible random samples of size nnn.

Part 1 done

2

Part 2 of 6: The sample mean keeps the population mean and shrinks the spread

Now compare Xˉ\bar XXˉ with the population. Its mean and spread follow exact rules, whatever the shape of the population.

For a random sample of size nnn, E(Xˉ)=μE(\bar X)=\muE(Xˉ)=μ and Var(Xˉ)=σ2n\mathrm{Var}(\bar X)=\frac{\sigma^2}{n}Var(Xˉ)=nσ2​. These follow from the rules for sums of independent variables.

The standard deviation of Xˉ\bar XXˉ is σn\frac{\sigma}{\sqrt n}n​σ​, called the standard error. It shrinks as nnn grows, so larger samples give means closer to μ\muμ.

Because E(Xˉ)=μE(\bar X)=\muE(Xˉ)=μ, sample means are not systematically too high or too low. They scatter evenly around the population mean.

Worked example

Finding a standard error

A population of masses has standard deviation σ=12 g\sigma=12\,\mathrm{g}σ=12g. Find the standard error of the mean for random samples of size n=36n=36n=36.
  1. 1
    Use standard error=σn\text{standard error}=\frac{\sigma}{\sqrt n}standard error=n​σ​.
  2. 2
    Substitute: 1236=126=2.0\frac{12}{\sqrt{36}}=\frac{12}{6}=2.036​12​=612​=2.0.

Answer: The standard error is 2.0 g2.0\,\mathrm{g}2.0g.

Key term

standard error

The standard error is the standard deviation of the sample mean, equal to σn\frac{\sigma}{\sqrt n}n​σ​ for random samples of size nnn.

Part 2 done

Quick check

One question on the part you just read

Easy

Identify the standard error of the sample mean for random samples of size nnn from a population with standard deviation σ\sigmaσ.
Show answer and explanation

Answer: A · Worked explanation

The variance of Xˉ\bar XXˉ is σ2n\frac{\sigma^2}{n}nσ2​, so its standard deviation is the square root, σn\frac{\sigma}{\sqrt n}n​σ​. The expression σn\frac{\sigma}{n}nσ​ forgets the square root, and σn\sigma\sqrt nσn​ multiplies instead of dividing, which would make the spread grow with nnn. The expression σ2n\frac{\sigma^2}{\sqrt n}n​σ2​ uses the variance in place of the standard deviation, so it is not in the units of the data.
3

Part 3 of 6: For large samples the means follow a normal distribution

The mean and spread are exact, but probabilities also need the shape. The central limit theorem supplies it.

Take a large random sample from any population with mean μ\muμ and finite variance σ2\sigma^2σ2. The theorem states that Xˉ\bar XXˉ is then approximately normal. In symbols, Xˉ≈N(μ,σ2n)\bar X\approx\mathrm{N}(\mu,\frac{\sigma^2}{n})Xˉ≈N(μ,nσ2​).

  • The observations form a random sample.
  • The observations are independent of each other.
  • The population has a finite mean and variance.
  • The sample size is large, usually n≥30n\geq30n≥30.

If the population is already normal, Xˉ\bar XXˉ is exactly normal for every nnn.

Common mistake

The theorem describes the distribution of sample means. It does not make the population, or one sample, normal.

Key term

central limit theorem

The central limit theorem states that the mean of a large random sample is approximately normal, with mean μ\muμ and variance σ2n\frac{\sigma^2}{n}nσ2​.

Part 3 done

4

Part 4 of 6: The theorem turns questions about means into normal probabilities

Now the theorem can be used. Find the standard error, then standardise with z=xˉ−μσ/nz=\frac{\bar x-\mu}{\sigma/\sqrt n}z=σ/n​xˉ−μ​ and use the normal distribution.

Worked example

Probability for a sample mean

Waiting times at a station have mean 8.0 min and standard deviation 6.0 min. The distribution is skewed. For a random sample of 36 waits, find the probability that the sample mean exceeds 9.5 min.
  1. 1
    Since n=36n=36n=36 is large, Xˉ\bar XXˉ is approximately normal with mean 8.0.
  2. 2
    Standard error: 6.036=1.0\frac{6.0}{\sqrt{36}}=1.036​6.0​=1.0, so Xˉ≈N(8.0, 1.02)\bar X\approx\mathrm{N}(8.0,\,1.0^2)Xˉ≈N(8.0,1.02).
  3. 3
    Standardise: z=9.5−8.01.0=1.5z=\frac{9.5-8.0}{1.0}=1.5z=1.09.5−8.0​=1.5.
  4. 4
    P(Xˉ>9.5)=P(Z>1.5)=1−0.9332=0.0668P(\bar X>9.5)=P(Z>1.5)=1-0.9332=0.0668P(Xˉ>9.5)=P(Z>1.5)=1−0.9332=0.0668.

Answer: The probability is ≈0.0668\approx0.0668≈0.0668.

A single wait varies far more than the mean of 36 waits. That is why the standard error, not σ\sigmaσ, is used here.

For a total rather than a mean, the total of nnn observations is approximately N(nμ, nσ2)\mathrm{N}(n\mu,\,n\sigma^2)N(nμ,nσ2). Choose the version that matches the question.

Exam tip

Use the standard error σn\frac{\sigma}{\sqrt n}n​σ​, not σ\sigmaσ, whenever a question is about a sample mean.

Part 4 done

Quick check

One question on the part you just read

Easy

State the approximate distribution of the sample mean Xˉ\bar XXˉ for a large random sample of size nnn from a population with mean μ\muμ and variance σ2\sigma^2σ2.
Show answer and explanation

Answer: C · Worked explanation

The central limit theorem gives a normal distribution with the same mean μ\muμ and variance σ2n\frac{\sigma^2}{n}nσ2​. The distribution N(μ,σ2)\mathrm{N}(\mu,\sigma^2)N(μ,σ2) would describe single observations from a normal population, not the means of samples. Sample means are less spread out, so their variance must be smaller.
5

Part 5 of 6: Larger samples give narrower sampling distributions

The standard error falls as nnn increases. To halve it, the sample size must be multiplied by four. Very large samples therefore give very precise sample means.

Standard error for a population with σ=12 g\sigma=12\,\mathrm{g}σ=12g

Sample size nnnStandard error (g)
94.0
362.0
1441.0
Sample size $n$
9
Standard error (g)
4.0
Sample size $n$
36
Standard error (g)
2.0
Sample size $n$
144
Standard error (g)
1.0

The graph shows two sampling distributions for a population with μ=50 g\mu=50\,\mathrm{g}μ=50g. Both are centred on 50, but the larger sample gives a taller, narrower curve.

Figure 1. Both curves are centred on 50 g, but the curve for n=36n=36n=36 is narrower and taller than the curve for n=9n=9n=9.

A narrower curve means a sample mean is more likely to land close to μ\muμ. This reduces random variation only. A biased sampling method stays biased at any size.

Part 5 done

6

Part 6 of 6: The sample size needed depends on the population shape

Now consider how large nnn must be. The usual guide is n≥30n\geq30n≥30, but this is only a rule of thumb.

A population that is close to symmetric needs a smaller sample. A strongly skewed population needs a larger one before the sampling distribution looks normal.

The theorem also needs independent observations. Sampling a large part of a small population breaks this, so the results become less reliable. Always check that the sample was random as well as large.

Note read

Check your recall[15 marks]

6 written questions. Answer each one in your head or on paper, then open the mark scheme.

  1. 1.State[4 marks]

    State the central limit theorem, including the conditions needed for it to apply.

    Show mark schemeHide mark scheme

    Model answer

    For a random sample of nnn independent observations from a population with mean μ\muμ and finite variance σ2\sigma^2σ2, where nnn is large, the sample mean Xˉ\bar XXˉ is approximately normal with mean μ\muμ and variance σ2n\frac{\sigma^2}{n}nσ2​.

    • 1 mark for stating that the sample is random and the observations are independent
    • 1 mark for stating that the population has a finite mean and variance, and nnn is large
    • 1 mark for stating that Xˉ\bar XXˉ is approximately normal
    • 1 mark for giving mean μ\muμ and variance σ2n\frac{\sigma^2}{n}nσ2​
  2. 2.Calculate[2 marks]

    A population has standard deviation 15. Calculate the standard error of the mean for random samples of size 25.

    Show mark schemeHide mark scheme

    Model answer

    Standard error =1525=155=3.0=\frac{15}{\sqrt{25}}=\frac{15}{5}=3.0=25​15​=515​=3.0.

    • 1 mark for using σn\frac{\sigma}{\sqrt n}n​σ​ with correct substitution
    • 1 mark for obtaining 3.0
  3. 3.Predict[2 marks]

    The sample size is multiplied by 9 and σ\sigmaσ is unchanged. Predict the effect on the standard error.

    Show mark schemeHide mark scheme

    Model answer

    The standard error is divided by 9=3\sqrt9=39​=3, so it becomes one third of its original value.

    • 1 mark for stating that the standard error decreases
    • 1 mark for stating that it is divided by 3
  4. 4.Explain[2 marks]

    Explain why the mean of a sample of 100 observations varies less than a single observation.

    Show mark schemeHide mark scheme

    Model answer

    The standard deviation of the sample mean is σ100=σ10\frac{\sigma}{\sqrt{100}}=\frac{\sigma}{10}100​σ​=10σ​. This is smaller than σ\sigmaσ, so sample means lie closer to μ\muμ than single observations do.

    • 1 mark for stating that the standard error is σn\frac{\sigma}{\sqrt n}n​σ​, which is smaller than σ\sigmaσ
    • 1 mark for linking this to sample means lying closer to μ\muμ
  5. 5.Determine[3 marks]

    A population has mean 40 and standard deviation 10. For a random sample of 64, determine the probability that the sample mean is less than 38.

    Show mark schemeHide mark scheme

    Model answer

    Standard error =1064=1.25=\frac{10}{\sqrt{64}}=1.25=64​10​=1.25. z=38−401.25=−1.6z=\frac{38-40}{1.25}=-1.6z=1.2538−40​=−1.6. P(Z<−1.6)≈0.0548P(Z<-1.6)\approx0.0548P(Z<−1.6)≈0.0548.

    • 1 mark for finding a standard error of 1.25
    • 1 mark for finding z=−1.6z=-1.6z=−1.6
    • 1 mark for obtaining a probability of ≈0.0548\approx0.0548≈0.0548
  6. 6.Distinguish[2 marks]

    Distinguish between the standard deviation σ\sigmaσ of a population and the standard error of the sample mean.

    Show mark schemeHide mark scheme

    Model answer

    The standard deviation σ\sigmaσ measures the spread of individual values in the population. The standard error σn\frac{\sigma}{\sqrt n}n​σ​ measures the spread of sample means and depends on nnn.

    • 1 mark for stating that σ\sigmaσ describes the spread of individual values
    • 1 mark for stating that the standard error describes the spread of sample means and depends on nnn

Practice this topic

10 quick-fire multiple-choice questions, easy to hard. Tap an answer to see if you're right and the worked explanation. (The written questions with mark schemes are under “Check your recall” above.)

0 right0 of 10 answered

Includes the 2 quick checks in the note above.

3.

Easy·Identify[1 mark]

Identify what the sampling distribution of the sample mean describes.
Show answer and explanation

Answer: B · Worked explanation

A sampling distribution describes the probability distribution of Xˉ\bar XXˉ over all possible random samples of one size. The spread of values within a single sample, or within the population, is a different thing. Sample size is fixed, not random.
4.

Medium·Calculate[1 mark]

A population has standard deviation 20. Calculate the standard error of the mean for random samples of size 25.
Show answer and explanation

Answer: D · Worked explanation

Standard error =2025=205=4.0=\frac{20}{\sqrt{25}}=\frac{20}{5}=4.0=25​20​=520​=4.0. The value 0.8 comes from dividing by nnn instead of n\sqrt nn​. The value 100 comes from multiplying by n\sqrt nn​, which would make the spread larger instead of smaller. The value 5.0 is simply 25\sqrt{25}25​, which does not use σ\sigmaσ at all.
5.

Medium·Determine[2 marks]

A population has mean 100 and standard deviation 15. For a random sample of 36, determine the probability that the sample mean exceeds 104.
Show answer and explanation

Answer: A · Worked explanation

The standard error is 1536=2.5\frac{15}{\sqrt{36}}=2.536​15​=2.5, so z=104−1002.5=1.6z=\frac{104-100}{2.5}=1.6z=2.5104−100​=1.6 and P(Z>1.6)≈0.0548P(Z>1.6)\approx0.0548P(Z>1.6)≈0.0548. The value 0.395 comes from using σ=15\sigma=15σ=15 instead of the standard error, which ignores the sample size. The value 0.945 is the lower tail, not the upper tail. Check which tail the question asks about before reading the calculator output.
6.

Medium·Identify[1 mark]

A population is strongly skewed. A random sample of 50 observations is taken. Identify the correct statement.
Show answer and explanation

Answer: B · Worked explanation

The central limit theorem applies to the sample mean for large nnn, here 50. It says nothing that changes the shape of the population or of the individual observations. The sample mean has variance σ250\frac{\sigma^2}{50}50σ2​, which is smaller than the population variance.
7.

Medium·Predict[1 mark]

The sample size is increased from 20 to 80, and σ\sigmaσ is unchanged. Predict what happens to the standard error.
Show answer and explanation

Answer: C · Worked explanation

The sample size is multiplied by 4, and the standard error is divided by 4=2\sqrt4=24​=2, so it is halved. Quartering would need the sample size to be multiplied by 16. The standard error does change, because it depends on nnn.
8.

Hard·Calculate[2 marks]

Items have mean mass 2.0 kg and standard deviation 0.5 kg. A random sample of 100 items is taken. Calculate the probability that the total mass exceeds 205 kg.
Show answer and explanation

Answer: B · Worked explanation

The total is approximately N(200, 100×0.25)=N(200,25)\mathrm{N}(200,\,100\times0.25)=\mathrm{N}(200,25)N(200,100×0.25)=N(200,25), with standard deviation 5. Then z=205−2005=1z=\frac{205-200}{5}=1z=5205−200​=1 and P(Z>1)≈0.159P(Z>1)\approx0.159P(Z>1)≈0.159. The value 0.240 uses a variance of 100×0.5=50100\times0.5=50100×0.5=50, multiplying σ\sigmaσ instead of σ2\sigma^2σ2 by nnn, which gives a standard deviation of 50\sqrt{50}50​. Variance, not standard deviation, is multiplied by nnn. The value 0.841 is the lower tail.
9.

Hard·Identify[2 marks]

Identify the situation in which the approximate normality of the sample mean is least justified.
Show answer and explanation

Answer: A · Worked explanation

A sample of 6 is too small for the theorem to give a good approximation, and the population is not normal. A sample of 45 is large enough, and a sample of 100 is large as well. A sample of 8 from a normal population gives an exactly normal mean.
10.

Hard·Determine[2 marks]

A population has standard deviation 14. For random samples of size 49, the central 95% of N(0,1)\mathrm{N}(0,1)N(0,1) lies between -1.96 and 1.96. Determine kkk such that P(μ−k<Xˉ<μ+k)=0.95P(\mu-k<\bar X<\mu+k)=0.95P(μ−k<Xˉ<μ+k)=0.95.
Show answer and explanation

Answer: D · Worked explanation

The standard error is 1449=2\frac{14}{\sqrt{49}}=249​14​=2, so k=1.96×2=3.92k=1.96\times2=3.92k=1.96×2=3.92. The value 27.4 uses σ=14\sigma=14σ=14 instead of the standard error. The value 0.56 divides by nnn instead of n\sqrt nn​. The value 1.96 is the zzz-value, with no conversion to the scale of Xˉ\bar XXˉ.

Keep going

Next topic4.16 Confidence intervalsPrevious topic4.14 Linear combinations of random variables

More in Topic 4 · Statistics and probability

  1. 4.11Chi-squared and t-testsNotes · 10 questions
  2. 4.12Designing data collectionNotes · 10 questions· HL only
  3. 4.13Non-linear regressionNotes · 10 questions· HL only
  4. 4.14Linear combinations of random variablesNotes · 10 questions· HL only
  5. 4.15Central limit theoremNotes · 10 questions· HL only
  6. 4.16Confidence intervalsNotes · 10 questions· HL only
  7. 4.17Poisson distributionNotes · 10 questions· HL only
  8. 4.18Hypothesis testingNotes · 10 questions· HL only
  9. 4.19Transition matrices and Markov chainsNotes · 10 questions· HL only

All IB Mathematics: Applications and Interpretation topics →

Draft: written by Scribe and checked automatically, not yet by a subject reviewer.

Practice questions are original, written by Scribe and checked automatically for correctness.

Topic codes follow the official numbering. Not endorsed by IB Diploma Programme or any exam board.

On this page

0/6 read

  1. 1Sample means vary from one sample to the next
  2. 2The sample mean keeps the population mean and shrinks the spread
  3. 3For large samples the means follow a normal distribution
  4. 4The theorem turns questions about means into normal probabilities
  5. 5Larger samples give narrower sampling distributions
  6. 6The sample size needed depends on the population shape
  7. Check your recall
  8. Practice: 10 questions
All coursesScribeHow it worksPricingPrivacyTerms

Notes and questions are written by Scribe in our own words. Topic codes follow the official course numbering so you can find your place. Not produced or endorsed by the IB, College Board, AQA, OCR, Pearson Edexcel or any exam board.

© 2026 Scribe

Sampling distributions for two sample sizes

Probability density (per g)404550556000.050.10.150.2Sample mean (g)
  • n = 9 (SE 4.0 g)
  • n = 36 (SE 2.0 g)