Tests for the difference between two Normal means
Studying this course? Keep your progress in one place.
Key facts
- : variances add.
- Under , .
- Hypotheses are about population means, not and .
- Two-tailed critical values (5%); one-tailed (5%) and (1%).
- With large samples, use in place of (Central Limit Theorem).
Distribution of the difference
It is Normal, with the means subtracted but the variances added.
Take independent samples from and . Then and .
The difference of independent Normal variables is Normal: means subtracted, variances added.
- Mean
- Variance
and independent. What is the distribution of ?
The test with known variances
Standardise with the combined standard error and compare with a Normal critical value.
Under the test statistic is with . Hypotheses are always about population means.
- 1
State hypotheses
and in terms of
- 2
Calculate
using the combined standard error
- 3
Compare
critical value, or find the -value
- 4
Conclude
in context
Worked example
, , ; , , . Find the test statistic.
- 1
Standard error .
- 2
Difference of means .
- 3
.
Which is a correct null hypothesis?
One-tailed and two-tailed tests
Choose from the wording: "differ" gives two-tailed, "greater" gives one-tailed.
"Differ" gives two-tailed , with the significance level split between both tails. "Greater" or "larger" gives one-tailed .
For a two-tailed test double the tail probability to get the -value. Reject if the significance level. Write the conclusion as evidence, never proof: "insufficient evidence that ...".
| Two-tailed | One-tailed | |
|---|---|---|
| Alternative | ||
| Critical value at 5% | 1.645 | |
| Critical value at 1% | 2.326 |
Two-tailed
- Alternative:
- Critical value at 5%:
- Critical value at 1%:
One-tailed
- Alternative:
- Critical value at 5%:
- 1.645
- Critical value at 1%:
- 2.326
A question asks whether the mean of is "greater than" the mean of at the level. The critical value is...
Unknown variances, large samples
For large samples, use and in place of the population variances.
When variances are unknown, replace and by the sample variances. For large samples the Central Limit Theorem says is approximately Normal even if the populations are not, and is close to .
Quote the Central Limit Theorem and the large samples when you do this.
- Variancesuse and
- Statistic
Worked example
, , ; , , . Test against at .
- 1
Standard error .
- 2
.
- 3
, so it is not in the critical region ().
Why can sample variances replace population variances in the test?
Try an exam question
Machine A fills bags with mean mass kg and machine B fills bags with mean mass kg. The population variances are known to be and . Test, at the level, whether the mean for machine A is greater than the mean for machine B.
[4 marks]
- [1], .
- [1]Standard error .
- [1].
- [1]: reject ; evidence that the mean for machine A is greater.
That's the notes covered.
Carry on to the next subtopic.
Try a question
Find the standard error of .
1 mark, marked against the mark scheme
Find the standard error of .
1 mark, marked against the mark scheme