Session 5 - Correlation

Question 1

Correlation Coefficient

Answer 1

Used to measure the strength of relationship between 2 continuous variables
Displayed using scatter diagrams - each point = one subject

Question 2

Calculation of Correlation Coefficient

Answer 2

Product of Differences from the Mean of 2 variables
For each observation - find difference from the mean, multiply the differences for the two variables, add them together
= ‘sum of products about the mean’

Question 3

Pearson’s Correlation coefficient

Answer 3

=r
= product moment correlation coefficient
Min -1, Max +1
No linear relationship r=0

Question 4

Correlation coefficient guidelines - rule of thumb

Answer 4

r = 0.0 No correlation
0.0 < r < 0.2 Very weak correlation
0.2 ≤ r < 0.4 Weak correlation
0.4 ≤ r < 0.6 Moderately strong correlation
0.6 ≤ r < 0.8 Strong correlation
0.8 ≤ r < 1.0 Very strong correlation
r = 1.0 Perfect correlation

Question 5

Assumptions for Pearson’s correlation coefficient

Answer 5

One or both variables are normally distributed (large deviations = unreliable p)
Observations are independent
Linear relationship
Two continuous variables

Question 6

What if assumption or normality not met for Pearson’s

Answer 6

Use alternative correlation coefficient
ie Spearman’s Correlation Coefficient
Works by first ranking observations
Then calculate product moment correlation of the ranks
Denoted as ρ or rs