Scribe logoScribeNotes & practice
Sign inStart free
  1. Notes
  2. ›IB Mathematics: Applications and Interpretation
  3. ›4 Statistics and probability
  4. ›4.16 Confidence intervals

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

4.16 Confidence intervals

Construct confidence intervals for a mean or a proportion, interpret them correctly, and explain how sample size and confidence level change their width.

  • 4 min read
  • 5 parts
  • 10 questions
  • Draft
  • HL only
Start a Quick 5 on this topic→Jump to the 10 questions
1/5A confidence interval turns one estimate into a range
  1. 1A confidence interval turns one estimate into a range
  2. 2Technology builds the interval for a mean from sample data
  3. 3A proportion has its own interval from the sample proportion
  4. 4A 95% interval describes the method, not one interval
  5. 5Larger samples narrow an interval and higher confidence widens it
  6. Check your recall
  7. Practice: 10 questions
1

Part 1 of 5: A confidence interval turns one estimate into a range

A school wants the mean time its students spend travelling each morning. It asks 40 students and finds a sample mean xˉ\bar{x}xˉ. This is a point estimate of the population mean μ\muμ.

A different sample of 40 would give a different xˉ\bar{x}xˉ. One number alone does not show how far it may lie from μ\muμ.

So we give a range instead, called a confidence interval. It has the form xˉ±E\bar{x}\pm Exˉ±E, where EEE is the margin of error. The interval is centred on the sample mean.

Common mistake

The sample mean is not the population mean. Samples from the same population usually give different sample means.

Key terms · tap one for its definition

point estimate

A point estimate is a single value calculated from sample data to estimate a population parameter, such as xˉ\bar{x}xˉ for μ\muμ.

confidence interval

A confidence interval is a range of values, built from sample data, that is likely to contain an unknown population parameter.

margin of error

The margin of error EEE is half the width of a confidence interval, added to and subtracted from the point estimate.

Part 1 done

2

Part 2 of 5: Technology builds the interval for a mean from sample data

Now we need a value for EEE. The margin is a critical value multiplied by the standard error. The standard error measures how much sample means vary between samples.

We estimate it by sn−1n\frac{s_{n-1}}{\sqrt{n}}n​sn−1​​, where sn−1s_{n-1}sn−1​ is the sample standard deviation. The critical value ttt comes from the t-distribution with n−1n-1n−1 degrees of freedom, a number set by the sample size. The interval is xˉ±tsn−1n\bar{x}\pm t\frac{s_{n-1}}{\sqrt{n}}xˉ±tn​sn−1​​.

A calculator’s t-interval function finds all of this from your data. If σ\sigmaσ is known, use the normal distribution, with zzz and σ\sigmaσ in place of ttt and sn−1s_{n-1}sn−1​.

Worked example

Estimate a mean battery life

A sample of 25 batteries has a mean life of 52.0 hours and sn−1=6.0s_{n-1}=6.0sn−1​=6.0 hours. Lives are normally distributed. Construct a 95% confidence interval for the population mean.
  1. 1
    Standard error: sn−1n=6.025=1.2\frac{s_{n-1}}{\sqrt{n}}=\frac{6.0}{\sqrt{25}}=1.2n​sn−1​​=25​6.0​=1.2 hours.
  2. 2
    Critical value: for 95% confidence and n−1=24n-1=24n−1=24 degrees of freedom, t≈2.064t\approx2.064t≈2.064.
  3. 3
    Margin of error: E=2.064×1.2≈2.48E=2.064\times1.2\approx2.48E=2.064×1.2≈2.48 hours.
  4. 4
    Interval: 52.0−2.48=49.5252.0-2.48=49.5252.0−2.48=49.52 to 52.0+2.48=54.4852.0+2.48=54.4852.0+2.48=54.48.

Answer: The 95% confidence interval is 49.5<μ<54.549.5<\mu<54.549.5<μ<54.5 hours.

Go deeperWhy the t-distribution is used

The t-distribution is bell-shaped like the normal distribution but has heavier tails. Using sn−1s_{n-1}sn−1​ in place of σ\sigmaσ adds extra uncertainty, so the multiplier must be larger.

As nnn grows, ttt approaches zzz. For 95% confidence, z=1.96z=1.96z=1.96, while t=2.064t=2.064t=2.064 when n=25n=25n=25.

Key terms · tap one for its definition

critical value

The critical value is the multiplier, taken from the normal or t-distribution, that sets the margin of error for a chosen confidence level.

standard error

The standard error of a sample mean is the standard deviation of its sampling variation, estimated by sn−1n\frac{s_{n-1}}{\sqrt{n}}n​sn−1​​.

Part 2 done

Quick check

One question on the part you just read

Easy

A sample of 30 students has a mean score of 64.2 and sn−1=8.0s_{n-1}=8.0sn−1​=8.0. State the point estimate of the population mean.
Show answer and explanation

Answer: A · Worked explanation

The sample mean xˉ=64.2\bar{x}=64.2xˉ=64.2 is the point estimate of the population mean. The value 8.0 is the sample standard deviation, and 1.46 is the standard error 8.030\frac{8.0}{\sqrt{30}}30​8.0​, which measures variation of the mean rather than estimating it. The point estimate is the sample statistic itself, not a measure of spread.
3

Part 3 of 5: A proportion has its own interval from the sample proportion

Sometimes the data are counts of successes, not measurements. Suppose 132 of 200 people polled say they cycle to work. The sample proportion is p^=132200=0.66\hat{p}=\frac{132}{200}=0.66p^​=200132​=0.66.

For a large sample, an approximate interval for the true proportion ppp is p^±zp^(1−p^)n\hat{p}\pm z\sqrt{\frac{\hat{p}(1-\hat{p})}{n}}p^​±znp^​(1−p^​)​​. The critical value zzz is 1.96 for 95% confidence.

The square root term is the standard error of the sample proportion. It is largest when p^\hat{p}p^​ is near 0.5, so such polls need larger samples.

Exam tip

State that np^n\hat{p}np^​ and n(1−p^)n(1-\hat{p})n(1−p^​) are both at least 10 before using this approximation. Here they are 132 and 68.

Worked example

Estimate a proportion

In a poll, 132 of 200 people say they cycle to work. Construct an approximate 95% confidence interval for the true proportion ppp.
  1. 1
    Standard error: 0.66×0.34200=0.001122≈0.0335\sqrt{\frac{0.66\times0.34}{200}}=\sqrt{0.001122}\approx0.03352000.66×0.34​​=0.001122​≈0.0335.
  2. 2
    Margin of error: E=1.96×0.0335≈0.0657E=1.96\times0.0335\approx0.0657E=1.96×0.0335≈0.0657.
  3. 3
    Lower end: 0.66−0.0657≈0.5940.66-0.0657\approx0.5940.66−0.0657≈0.594. Upper end: 0.66+0.0657≈0.7260.66+0.0657\approx0.7260.66+0.0657≈0.726.

Answer: The 95% confidence interval is 0.594<p<0.7260.594<p<0.7260.594<p<0.726.

Key term

sample proportion

The sample proportion p^\hat{p}p^​ is the number of successes in a sample divided by the sample size nnn.

Part 3 done

4

Part 4 of 5: A 95% interval describes the method, not one interval

Now that we can calculate an interval, we must read it correctly. The confidence level says how often the method succeeds. Take many samples and build a 95% interval from each. About 95% of those intervals would contain μ\muμ.

Any one interval either contains μ\muμ or it does not. So we say we are 95% confident that μ\muμ lies in it. For the batteries, we are 95% confident that the mean life is between 49.5 and 54.5 hours.

Common mistake

A 95% confidence interval does not contain 95% of the data values. It is a statement about the population mean.

Key term

confidence level

The confidence level is the long-run percentage of intervals, built by the same method from repeated samples, that contain the parameter.

Part 4 done

Quick check

One question on the part you just read

Easy

A confidence interval is written xˉ±E\bar{x}\pm Exˉ±E. Identify what EEE represents.
Show answer and explanation

Answer: A · Worked explanation

The quantity EEE is the margin of error. It is added to and subtracted from the sample mean to give the two ends of the interval, so it is half the interval’s width. The sample standard deviation describes the spread of the data, and it is only one ingredient in finding EEE through the standard error.
5

Part 5 of 5: Larger samples narrow an interval and higher confidence widens it

Two choices control how wide an interval is. Both act through the margin of error.

  • A larger sample size nnn reduces the standard error, so the interval narrows.
  • A higher confidence level needs a larger critical value, so the interval widens.

The figure uses the battery sample at three confidence levels. The 99% interval is the most likely to capture μ\muμ but the least precise.

Figure 1. For the same sample, the 99% interval is wider than the 95% interval, which is wider than the 90% interval.

Quadrupling nnn roughly halves the width, since the standard error depends on n\sqrt{n}n​. Precision therefore costs a larger sample.

Narrower intervals are more precise, but a narrow interval at low confidence can mislead. The right balance depends on the cost of being wrong.

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.Define[2 marks]

    Define the term confidence level.

    Show mark schemeHide mark scheme

    Model answer

    The confidence level is the percentage of intervals, built by the same method from repeated samples, that would contain the population parameter.

    • 1 mark for stating that it is a percentage of intervals from repeated samples built by the same method
    • 1 mark for stating that these intervals contain the population parameter
  2. 2.Calculate[3 marks]

    A sample of 36 plants has a mean height of 40.0 cm and sn−1=9.0s_{n-1}=9.0sn−1​=9.0 cm. Heights are normally distributed. Using a critical value t=2.03t=2.03t=2.03, calculate a 95% confidence interval for the population mean.

    Show mark schemeHide mark scheme

    Model answer

    Standard error =9.036=1.5=\frac{9.0}{\sqrt{36}}=1.5=36​9.0​=1.5. Margin E=2.03×1.5≈3.05E=2.03\times1.5\approx3.05E=2.03×1.5≈3.05. The interval is 37.0<μ<43.037.0<\mu<43.037.0<μ<43.0 cm.

    • 1 mark for calculating the standard error as 1.5
    • 1 mark for calculating the margin of error as about 3.05
    • 1 mark for giving the interval as 37.0 to 43.0 cm
  3. 3.Calculate[3 marks]

    In a sample of 150 voters, 90 support a plan. Using z=1.96z=1.96z=1.96, calculate an approximate 95% confidence interval for the true proportion.

    Show mark schemeHide mark scheme

    Model answer

    p^=0.6\hat{p}=0.6p^​=0.6. Standard error =0.6×0.4150=0.04=\sqrt{\frac{0.6\times0.4}{150}}=0.04=1500.6×0.4​​=0.04. Margin =1.96×0.04≈0.0784=1.96\times0.04\approx0.0784=1.96×0.04≈0.0784. The interval is 0.522<p<0.6780.522<p<0.6780.522<p<0.678.

    • 1 mark for finding p^=0.6\hat{p}=0.6p^​=0.6
    • 1 mark for finding the standard error as 0.04 and the margin as about 0.0784
    • 1 mark for giving the interval as 0.522 to 0.678
  4. 4.Explain[2 marks]

    Explain why a 99% confidence interval is wider than a 95% interval from the same sample.

    Show mark schemeHide mark scheme

    Model answer

    A higher confidence level needs a larger critical value. The margin of error is the critical value multiplied by the standard error, so it increases.

    • 1 mark for stating that a higher confidence level gives a larger critical value
    • 1 mark for linking the margin of error to the critical value, so the interval is wider
  5. 5.Deduce[2 marks]

    A 95% confidence interval for a mean is 49.5 to 54.5 hours. Deduce the sample mean and the margin of error.

    Show mark schemeHide mark scheme

    Model answer

    The interval is centred on the sample mean, so xˉ=49.5+54.52=52.0\bar{x}=\frac{49.5+54.5}{2}=52.0xˉ=249.5+54.5​=52.0 hours. The margin of error is 54.5−52.0=2.554.5-52.0=2.554.5−52.0=2.5 hours.

    • 1 mark for finding the sample mean as 52.0 hours
    • 1 mark for finding the margin of error as 2.5 hours
  6. 6.Discuss[3 marks]

    Discuss the statement “there is a 95% probability that the population mean lies between 49.5 and 54.5”.

    Show mark schemeHide mark scheme

    Model answer

    The statement is not correct. The population mean is a fixed value, so this interval either contains it or does not. The 95% refers to the method: about 95% of such intervals would contain the mean, so we are 95% confident.

    • 1 mark for stating that the population mean is a fixed value
    • 1 mark for stating that the interval either contains the mean or does not
    • 1 mark for stating that 95% refers to the method or that we can only be 95% confident

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 the change that makes a confidence interval for a mean wider, with all else unchanged.
Show answer and explanation

Answer: A · Worked explanation

A higher confidence level needs a larger critical value, so the margin of error grows. A larger sample or a smaller standard deviation reduces the standard error and narrows the interval, and a lower confidence level also narrows it.
4.

Medium·Calculate[2 marks]

A normal population has known standard deviation σ=12\sigma=12σ=12. A sample of n=36n=36n=36 has mean 80. Using z=1.96z=1.96z=1.96, calculate a 95% confidence interval for the population mean.
Show answer and explanation

Answer: A · Worked explanation

The standard error is 1236=2\frac{12}{\sqrt{36}}=236​12​=2, so the margin of error is E=1.96×2=3.92E=1.96\times2=3.92E=1.96×2=3.92. The interval is 80±3.9280\pm3.9280±3.92, which is 76.08 to 83.92, or 76.1 to 83.9. The range 56.5 to 103.5 comes from multiplying 1.96 by σ=12\sigma=12σ=12 directly, instead of by the standard error of the mean.
5.

Medium·Identify[2 marks]

A 95% confidence interval for the mean mass of apples is 148 g to 156 g. Identify the conclusion that is justified.
Show answer and explanation

Answer: A · Worked explanation

The interval is a statement about the population mean. The claim about 95% of apples confuses the interval with the spread of the data. The sample mean, 152 g, is known and lies in the interval for certain.
6.

Medium·Predict[2 marks]

A 95% confidence interval for a mean, from 25 observations, has width 8.0. A new sample of 100 observations comes from the same population with a similar standard deviation. Predict the approximate width of the new 95% interval.
Show answer and explanation

Answer: A · Worked explanation

The standard error depends on n\sqrt{n}n​, so quadrupling nnn from 25 to 100 roughly halves it, and the width falls from 8.0 to about 4.0. The critical value also falls slightly, so the width is a little under 4.0. The value 2.0 comes from assuming width is inversely proportional to nnn instead of n\sqrt{n}n​.
7.

Medium·Calculate[2 marks]

In a sample of 80 people, 56 own a bicycle. Using z=1.96z=1.96z=1.96, calculate an approximate 95% confidence interval for the true proportion.
Show answer and explanation

Answer: A · Worked explanation

Here p^=5680=0.7\hat{p}=\frac{56}{80}=0.7p^​=8056​=0.7 and the standard error is 0.7×0.380≈0.0512\sqrt{\frac{0.7\times0.3}{80}}\approx0.0512800.7×0.3​​≈0.0512. Then E=1.96×0.0512≈0.100E=1.96\times0.0512\approx0.100E=1.96×0.0512≈0.100, so the interval is 0.600 to 0.800. The interval 0.616 to 0.784 uses the critical value 1.645, which belongs to 90% confidence, not 95%. Always use the critical value that matches the confidence level.
8.

Hard·Determine[2 marks]

A 95% confidence interval from n=25n=25n=25 observations is 49.5 to 54.5 hours, using t=2.064t=2.064t=2.064. Determine the sample standard deviation sn−1s_{n-1}sn−1​.
Show answer and explanation

Answer: A · Worked explanation

The margin of error is half the width, so E=54.5−49.52=2.5E=\frac{54.5-49.5}{2}=2.5E=254.5−49.5​=2.5. Then sn−125=2.52.064≈1.211\frac{s_{n-1}}{\sqrt{25}}=\frac{2.5}{2.064}\approx1.21125​sn−1​​=2.0642.5​≈1.211, so sn−1≈6.06s_{n-1}\approx6.06sn−1​≈6.06 hours. The value 12.5 comes from ignoring the critical value, and 1.21 is only the standard error, not the standard deviation. The margin of error must be half the width before it is used.
9.

Hard·Deduce[2 marks]

Two researchers take separate samples from one population. Their 95% intervals for μ\muμ are 10.2 to 12.8 and 11.5 to 13.1. Deduce which conclusion is correct.
Show answer and explanation

Answer: A · Worked explanation

Different samples give different intervals, so two valid intervals need not match. Both may contain a value such as 12, so both can contain μ\muμ. The method captures μ\muμ about 95% of the time, but any one interval could be among those that miss, so the overlap cannot be assumed to contain the mean.
10.

Hard·Compare[2 marks]

Interval X uses n=50n=50n=50 at 95% confidence. Interval Y uses n=200n=200n=200 at 99% confidence. Both samples have the same standard deviation. Compare the widths using normal critical values 1.96 and 2.576.
Show answer and explanation

Answer: A · Worked explanation

The width is proportional to critical value divided by n\sqrt{n}n​. For X this is 1.9650≈0.277\frac{1.96}{\sqrt{50}}\approx0.27750​1.96​≈0.277 and for Y it is 2.576200≈0.182\frac{2.576}{\sqrt{200}}\approx0.182200​2.576​≈0.182, so Y is about 66% as wide. It is not exactly half, because the critical value also rose.

Keep going

Next topic4.17 Poisson distributionPrevious topic4.15 Central limit theorem

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

0/5 read

  1. 1A confidence interval turns one estimate into a range
  2. 2Technology builds the interval for a mean from sample data
  3. 3A proportion has its own interval from the sample proportion
  4. 4A 95% interval describes the method, not one interval
  5. 5Larger samples narrow an interval and higher confidence widens it
  6. Check your recall
  7. 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