Hypothesis tests for correlation
Studying this course? Keep your progress in one place.
Key facts
- The product moment correlation coefficient lies between and and measures linear association only.
- Hypotheses use the population coefficient : .
- Reject if is beyond the critical value for the sample size, level and tail.
- With a -value: reject if is less than the significance level.
- Correlation does not imply causation; conclude in context.
Correlation coefficients
measures how close points lie to a straight line: close to is strong, close to is weak.
The product moment correlation coefficient measures how close the points on a scatter diagram lie to a straight line, with . Close to is strong positive correlation, close to strong negative, and close to little or no linear correlation.
You do not calculate in this topic: it is given. It measures linear association only, so does not rule out a non-linear relationship.
Which value of shows the strongest linear correlation?
Hypotheses about the population
; is one-tailed ( or ) if the direction is known, else .
A sample varies from sample to sample, so we test a claim about the population correlation coefficient . The null hypothesis is that there is no correlation: .
The alternative is one-tailed if the direction is known in advance, or , and two-tailed if any correlation is of interest, . Choose the direction from the question wording, before looking at . The test assumes a random sample.
- one-tailed or
- two-tailed
A teacher wants to test whether revision time is positively correlated with test score. Which is ?
Critical value method
Compare with the critical value for the sample size, significance level and type of test.
A table gives the critical value of for each sample size , significance level and type of test.
- one-tailed upper test: reject if critical value;
- one-tailed lower test: reject if critical value;
- two-tailed test: reject if critical value (use the two-tailed table).
Use the number of pairs for , and check you are using the right table.
Worked example
, , at the level. The critical value is . Carry out the test.
- 1
, .
- 2
Critical value for , one-tailed: .
- 3
, so is in the critical region.
, at , and . The critical value is . What is the decision?
The p-value method
Compare the -value with the significance level: smaller means reject .
If a p-value is given, it is the probability of a sample correlation at least this extreme if . Compare it with the significance level (halving the level for a two-tailed test if the -value is for one tail). The smaller the -value, the stronger the evidence against .
- significance levelreject
- significance leveldo not reject
Worked example
For with and , the -value is . What is the decision at and at ?
- 1
At : , so do not reject .
- 2
At : , so reject .
at the level (one-tailed). What is the decision?
Conclusions and cautions
State the decision in context, and add cautions: correlation is not causation, and is only linear.
State the decision and meaning in context: 'there is sufficient evidence, at the level, of positive correlation between hours revising and test score'. If is not rejected, say there is insufficient evidence of correlation. Do not say is proved.
Cautions: correlation does not imply causation (another variable may explain both); measures only a linear relationship; extrapolation beyond the data is unreliable; a one-tailed test is only valid if the direction was decided before seeing the data.
- 1
Hypotheses
H₀: ρ = 0 and H₁ from the question
- 2
Find critical value
from n, the level and the type of test
- 3
Compare
sample r against the critical value
- 4
Decide
reject or do not reject H₀
- 5
Conclude
in context, with a caution if needed
A test finds significant positive correlation between ice cream sales and drownings. What is the best comment?
Try an exam question
A teacher collects pairs of data on hours of revision and test score and finds . The critical value for a one-tailed test with is . Test whether there is positive correlation, and state your conclusion in context.
[4 marks]
- [1],
- [1]Compare with the critical value
- [1], so reject
- [1]There is sufficient evidence, at the level, of positive correlation between revision hours and test score
That's the notes covered.
Carry on to the next subtopic.
Try a question
Which pair of hypotheses is correct for the teacher's test?
1 mark, marked against the mark scheme
Which pair of hypotheses is correct for the teacher's test?
1 mark, marked against the mark scheme