Confidence intervals from small samples
Studying this course? Keep your progress in one place.
Key facts
- For a small sample from a normal population with unknown variance, .
- The interval is with degrees of freedom.
- For a interval read the column for tail probability .
- critical values are larger than values, so small-sample intervals are wider.
- Check whether a claimed mean lies inside the interval, and answer in context.
Why the t-distribution is needed
With a small sample, can be well away from , so the standardised mean follows a -distribution with heavier tails.
For a normal population with unknown variance, , the t-distribution with degrees of freedom.
It is symmetric about with heavier tails than the standard normal, so its critical values are larger. As grows it approaches the normal distribution, which is why the normal can be used for large samples.
Compared with the standard normal, the t-distribution has
The interval
Go either side of by a critical value times the standard error .
The critical value is the value from the table with and tail probability .
For : uses the column () and uses the column ().
Find and first, using the unbiased variance. The standard error is .
- Interval for
- Unbiased variance
For a interval, which tail probability column do you use?
A small-sample interval
Using instead of widens the interval to allow for the small sample.
Chocolate bars: , , . The margin of error is , so the interval is margin of error.
The number line compares the interval with the too-narrow interval from .
Worked example
Chocolate bars: , , . Find a confidence interval for the mean mass.
- 1
, and the critical value is .
- 2
Standard error .
- 3
Margin .
- 4
Interval .
For , which critical value gives a interval?
Width, assumptions and inference
The interval is wider for higher confidence, larger or smaller , and it assumes a normal population and a random sample.
A smaller also means a larger value. The method assumes a normal population, which matters most when is small.
If a claimed mean lies inside the interval, the data are consistent with the claim. If it lies outside, there is evidence at the matching level that the mean is different, and the interval shows whether it is higher or lower. Always answer in context.
Wider interval
- Higher confidence level
- Larger
- Smaller (and larger )
Narrower interval
- Lower confidence level
- Smaller
- Larger
A interval for a mean is . A company claims the mean is . What do you conclude?
Try an exam question
The times, in seconds, taken by a random sample of pupils to complete a puzzle have mean and unbiased standard deviation . Assuming times are normally distributed, find a confidence interval for the mean time. (For the critical value is .)
[4 marks]
- [1] and .
- [1]Standard error .
- [1]Interval .
- [1].
That's the notes covered.
Carry on to the next subtopic.
Try a question
State the number of degrees of freedom of the -distribution used to construct a confidence interval for the population mean.
1 mark, marked against the mark scheme
State the number of degrees of freedom of the -distribution used to construct a confidence interval for the population mean.
1 mark, marked against the mark scheme