Confidence intervals for a Normal mean

Confidence intervals for a Normal mean

Spec 5.2–3 · Edexcel A-Level Further Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • A confidence interval is a range, from a sample, likely to contain the population parameter.
  • Say "we are 95%95\% confident that μ\mu lies in the interval", never "a 95%95\% probability".
  • With σ\sigma known: xˉ±zσn\bar x\pm z\frac{\sigma}{\sqrt n}, with z=1.645z=1.645, 1.961.96, 2.5762.576 for 90%90\%, 95%95\%, 99%99\%.
  • Width is 2zσn2z\frac{\sigma}{\sqrt n}: four times the sample halves it. Round nn up.
  • A (100−α)%(100-\alpha)\% interval matches a two-tailed test at the α%\alpha\% level.

What an interval means

The confidence level describes the method over repeated samples, not one calculated interval.

A confidence interval is a range of values, calculated from a sample, that is likely to contain the unknown population parameter.

A 95%95\% interval is built by a method that, over many samples, gives intervals containing the true μ\mu about 95%95\% of the time. Once calculated, μ\mu is a fixed number that is inside it or not. So say: "we are 95%95\% confident that μ\mu lies in the interval".

misses μμ (fixed)
Illustrative: five 95% intervals from repeated samples. Each is calculated by the same method; most capture the fixed true mean μ, but this one does not.
  • 95% confidencethe method captures μ\mu in 95% of repeated samples

What does '95%95\% confidence' describe?

Interval with known variance

Use xˉ±zσn\bar x\pm z\frac{\sigma}{\sqrt n}, with zz from the confidence level.

If X∼N(μ,σ2)X\sim\mathrm{N}(\mu,\sigma^2) with σ\sigma known, then Xˉ∼N(μ,σ2n)\bar X\sim\mathrm{N}\left(\mu,\frac{\sigma^2}{n}\right), so the standard error is σn\frac{\sigma}{\sqrt n}.

The interval is xˉ±zσn\bar x\pm z\frac{\sigma}{\sqrt n}. Check the level against the tails: 95%95\% leaves 2.5%2.5\% in each tail, so z=1.96z=1.96.

  • Confidence intervalxˉ±z σn\bar x\pm z\,\dfrac{\sigma}{\sqrt n}

Worked example

σ=6\sigma=6, n=25n=25, xˉ=503.2\bar x=503.2. Find a 95%95\% confidence interval for μ\mu.

499500501502503504505506507500.85503.2505.55

90%

  • z = 1.645

95%

  • z = 1.96

99%

  • z = 2.576

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

Width and sample size

The width is 2zσn2z\frac{\sigma}{\sqrt n}: it falls as nn grows, and you must round nn up.

The interval is symmetric about xˉ\bar x, with width 2zσn2z\frac{\sigma}{\sqrt n}.

Width falls as nn increases (halving it needs four times the sample size) and rises with the confidence level or with σ\sigma. For a maximum width ww, solve 2zσn≤w2z\frac{\sigma}{\sqrt n}\le w, so n≥(2zσw)2n\ge\left(\frac{2z\sigma}{w}\right)^2, then round up.

102030405060708090100510152025303540xy(25, 7.84)(100, 3.92)width
Illustrative: 95% interval width 2 × 1.96 × 10 ÷ √n. Quadrupling n from 25 to 100 halves the width.
  • Width2zσn2z\dfrac{\sigma}{\sqrt n}
  • Sample sizen≥(2zσw)2n\ge\left(\dfrac{2z\sigma}{w}\right)^2, round up

Worked example

σ=2.4\sigma=2.4, 90%90\% confidence, maximum width w=0.8w=0.8. Find the smallest nn.

The sample size is increased. The confidence interval width...

Intervals and hypothesis tests

If the claimed mean lies outside the interval, a matching two-tailed test rejects it.

A (100−α)%(100-\alpha)\% confidence interval contains the values of μ\mu that a two-tailed test at the α%\alpha\% level would not reject. So 95%95\% matches 5%5\%, and 99%99\% matches 1%1\%.

If a claimed value μ0\mu_0 lies outside the interval, reject H0:μ=μ0H_0:\mu=\mu_0; if inside, there is no evidence against it. This agrees with z=xˉ−μ0σ/nz=\frac{\bar x-\mu_0}{\sigma/\sqrt n} against the critical value.

12.412.512.612.712.812.91313.113.212.612.8413.0

Worked example

xˉ=12.84\bar x=12.84, σ=0.5\sigma=0.5, n=40n=40. Use a 99%99\% interval to test μ=13.0\mu=13.0 and μ=12.6\mu=12.6.

A 95%95\% interval for μ\mu is (4.1, 5.3)(4.1,\,5.3). Is H0:μ=5.5H_0:\mu=5.5 rejected at the 5%5\% level (two-tailed)?

Try an exam question

The masses of bags of flour are Normally distributed with standard deviation 4.54.5 g. A random sample of 3636 bags has mean mass 28.428.4 g. Find a 95%95\% confidence interval for the mean mass, and say what it shows about a claim that the mean is 3030 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