Confidence intervals for a mean

Confidence intervals for a mean

Spec SH1–SH3 · AQA A-Level Further Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • The sample mean of a normal population is normal: Xˉ∼N(μ,σ2n)\bar X\sim N\left(\mu,\dfrac{\sigma^2}{n}\right).
  • The standard error σn\dfrac{\sigma}{\sqrt n} shrinks as nn grows.
  • A C%C\% interval for μ\mu is xˉ±zσn\bar x\pm z\dfrac{\sigma}{\sqrt n}, with z=1.645z=1.645, 1.961.96 or 2.5762.576 for 90%90\%, 95%95\% or 99%99\%.
  • With a large sample and unknown σ\sigma, use ss (divisor n−1n-1) in its place.
  • Wider intervals come from higher confidence, larger spread or smaller nn.

Distribution of the sample mean

For a normal population, the sample mean is also normal, with the same mean and a variance divided by nn.

If a population is normal with mean μ\mu and variance σ2\sigma^2, the mean Xˉ\bar X of a random sample of size nn is also normal.

The standard deviation of Xˉ\bar X is the standard error. It shrinks as nn grows, so larger samples pin down μ\mu more precisely.

A confidence interval uses this to give a range of plausible values for μ\mu.

203040506070800.050.10.150.2xyPopulationSample mean, n = 25
Illustrative: a population N(50, 10²) and the distribution of the mean of samples of size 25, N(50, 2²). The sample mean is much less spread.
  • Sample meanXˉ∼N(μ, σ2n)\bar X\sim N\left(\mu,\ \dfrac{\sigma^2}{n}\right)
  • Standard errorσn\dfrac{\sigma}{\sqrt n}

A population has σ=12\sigma=12. What is the standard error of the mean for a sample of size n=36n=36?

Interval when the variance is known

Go either side of the sample mean by zz standard errors, where zz is chosen from the confidence level.

A symmetric C%C\% confidence interval for μ\mu is xˉ±zσn\bar x\pm z\dfrac{\sigma}{\sqrt n}.

The critical value zz has C%C\% of the standard normal distribution between −z-z and zz. The interval is a symmetric interval, centred on the sample mean.

  • 90%90\%: z=1.645z=1.645
  • 95%95\%: z=1.96z=1.96
  • 99%99\%: z=2.576z=2.576
−4−3−2−112340.10.20.30.4xy−1.961.96standard normal
The shaded area between −1.96 and 1.96 is 95% of the distribution.

Worked example

A population has σ=12\sigma=12. A sample of n=36n=36 has mean xˉ=504\bar x=504. Find a 95%95\% confidence interval for μ\mu.

Which zz value is used for a 99%99\% confidence interval?

Unknown variance, large sample

With a large sample, use the sample estimate ss in place of σ\sigma and keep the same zz values.

If the sample is large (roughly n≥30n\ge30), ss is a reliable estimate of σ\sigma, so the interval is xˉ±zsn\bar x\pm z\dfrac{s}{\sqrt n}.

Use the unbiased estimate of the variance, with divisor n−1n-1.

−4−3−2−112340.10.20.30.4xyz = −1.96z = 1.96Standard normal
The central 95% of a standard normal distribution lies between z = −1.96 and z = 1.96.
  • Large-sample intervalxˉ±zsn\bar x\pm z\dfrac{s}{\sqrt n}
  • Unbiased variances2=∑x2−nxˉ2n−1s^2=\dfrac{\sum x^2-n\bar x^2}{n-1}

Worked example

A sample has n=50n=50, xˉ=1172\bar x=1172 and s=96s=96. Find a 95%95\% confidence interval for μ\mu.

Why can we use ss in place of σ\sigma and still use zz values?

Making inferences from an interval

The interval shows which values of μ\mu are plausible, and its width depends on confidence level, spread and sample size.

If a claimed value of μ\mu lies inside the interval, the data are consistent with the claim. If it lies outside, there is evidence at the matching level (for example 5%5\% for a 95%95\% interval) that the mean is different, and the interval shows in which direction. Always answer in context.

For a 95%95\% interval of width at most ww, solve 2×1.96×σn≤w2\times1.96\times\frac{\sigma}{\sqrt n}\le w for nn and round up.

44454647484950515253545556575847x̄ = 5053claim: 55
Illustrative: a 95% confidence interval (47, 53) around a sample mean of 50. A claimed mean of 55 lies outside it, so there is evidence it is different.

Wider interval

  • Higher confidence level (larger zz)
  • Larger σ\sigma
  • Smaller nn

Narrower interval

  • Lower confidence level
  • Smaller σ\sigma
  • Larger nn

Which change makes a confidence interval narrower?

Try an exam question

The masses of packets of sugar are normally distributed with standard deviation 66 g. A random sample of 2525 packets has mean mass 502502 g. Find a 95%95\% confidence interval for the mean mass, and say whether it is consistent with the manufacturer's claim that the mean is 500500 g.

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Find the standard error of the sample mean.

1 mark, marked against the mark scheme

Revision notes