Hypothesis tests for the difference between two means
Studying this course? Keep your progress in one place.
Key facts
- against about the population means.
- : means subtract but variances add.
- Under : .
- With unknown variances and large samples, use and .
Comparing two means
Two independent samples test whether the population means differ.
Two independent random samples are taken, one from each population. has mean , variance , sample size and sample mean ; similarly for . The statistic tested is the difference of the sample means, .
The null hypothesis holds the equality, and the hypotheses concern population means, not sample means.
- H₀
- H₁, or
Which hypotheses are correct for comparing two population means?
The distribution of the difference
The means subtract but the variances add.
If and are independent, then and .
Variances add, because subtracting independent random variables does not remove any variability. The standard error of the difference is .
- X̄ − Ȳ
Worked example
, , , . Find the variance of .
- 1
.
- 2
.
- 3
Variances add.
If and , then is:
Carrying out the test
Under the statistic is ; compare it with the critical value.
Under (), the test statistic is .
Compare with the critical value: (two-tailed ), (one-tailed ), (one-tailed ), (two-tailed ).
Worked example
, , , , , . Test against at .
- 1
Variance of the difference: .
- 2
Standard error: .
- 3
.
- 4
Critical value is and .
In the fertiliser test, and the critical value is . What is the conclusion?
Large samples, unknown variances
With both samples large, use the sample variances and .
Use the unbiased estimates and . If both samples are large, the Central Limit Theorem makes and approximately Normal, and .
Example: , , , , , . The variance is and , so insufficient evidence of a difference at .
- Large samples
Two large samples, variances unknown. What replaces in the standard error?
Conclusions and assumptions
Conclude in context at the stated level, and state the assumptions made.
Give the conclusion in context: "there is evidence that method gives a shorter mean time than method ", or "there is insufficient evidence that the mean masses differ". Never claim is proved.
A test about means says nothing about every individual: some members of the population with the smaller mean may still exceed some with the larger mean. A result can be significant at but not at .
Small samples
- Populations Normal
- Variances known
- Samples random and independent
Large samples
- Central Limit Theorem used
- Sample variances may replace σ²
- Samples random and independent
in a one-tailed (lower) test. The result is significant at:
Try an exam question
Machine fills tins with standard deviation g, and machine with standard deviation g. A random sample of tins from has mean g and a random sample of tins from has mean g. Test, at the level, whether the means of the two machines differ.
[5 marks]
- [1], .
- [1]Variance of the difference .
- [1].
- [1]Critical values are , so do not reject .
- [1]There is insufficient evidence that the mean fill of the two machines differs.
That's the notes covered.
Carry on to the next subtopic.
Try a question
Under , what is the variance of ?
1 mark, marked against the mark scheme
Under , what is the variance of ?
1 mark, marked against the mark scheme