Week 10: Correlation And Regression

Question 1

Correlation types:

What type of correlation coefficient would be used to examine the relationship between two continuous variables?

Answer 1

Pearsons product-moment correlation coefficient (r)

Question 2

Correlation types

What correlation coefficient would be use to examine the relationship between two ordinal variables?

Answer 2

Spearman rank correlation (rs)
Eg relationship between height (very short, short, tall) and weight (underweight, average, overweight, obese)
Ordinal vs ordinal

Question 3

What correlation coefficient would be used to examine the relationship between one continuous variable and one dichotomous variable?
Eg eg relationship between trunk muscle strength (endurance holding time) and future traumatic knee injury

Answer 3

Point biserial correlation coefficient (r pb)

Question 4

What correlation coefficient would be usd to examine the relationship between 2 dichotomous variables (X and Y)
Eg eg in basketball players what is the relationship between a history of low back pain (yes/no) and future knee injury (yes/no)

Answer 4

Phi coefficient

Question 5

What is the function of scatterplots?

Answer 5

Shows the form (linear/ non linear), direction (pos/neg) and strength (correlation coefficient)

• it may be linear but can also be curvilinear or non linear
• a correlation weakens after a certain point when data is curvilinear
• correlation analysis is only appropriate with linear data because with curvilinear relationship would be underestimated.

Question 6

What is the correlation coefficient squared for?

Answer 6

Denotes how much of the variation in one variable can be explained by the other.
Eg a study on the relationship between the amount lifted at work and the incidence of low back pain reported r^2 as 0.65: 65% of stories can be explained by degree of lifting at work
-other factors are responsible for the remaining 35%

Question 7

What is Regression?

What is R^2? (Regression coefficient)

Answer 7

Regression is a statistical technique for estimating the relationship among variables. Ie how much the dependant variable can be explained by the independent variables.
Regression coefficient tells us how much of the variation in one variable can be explained by the other variable.