Correlation

Question 1

What is correlation?

Answer 1

A way of measuring the extent to which two distributions of scores are related

Question 2

What is the dependent variable?

Answer 2

The response outcome variable - we expect to observe a change

Question 3

What is the independent variable?

Answer 3

The variable we expect to create a change in the value of the DV

Question 4

What research design is needed to determine cause and effect?

Answer 4

Experimental study = direct and controlled manipulation of IV

Question 5

What research design can be used if experimental ones are ethical?

Answer 5

Observationial = looking at behaviour to establish some sort of relationship

Question 6

What is a scatterplot?

Answer 6

Helps to understand the relationship between variables distribution by graphing pairs together

Question 7

What are patterns of linear relationships?

Answer 7

Pos: as scores on IV go up, so do DV
Neg: as scores on IV go up, DV go down

Question 8

What is a measure to quantify the relationship between variables?

Answer 8

Using Pearsons correlation coefficient (r) - measures the extent to which variables are related:
r = ZxZy / N

Question 9

What do the points of r reveal about the relationship?

Answer 9

Perfect pos: +1
Perfect neg: -1
No relationship: 0

Question 10

What is the link between z scores and correlation coefficient?

Answer 10

Z scores reveal whether a score (X or Y) is above (+Z) or below (-Z) a mean. By multiplying z values for each person = cross products

Question 11

What are 3 possibilities of causality and what does this mean?

Answer 11

1. X causing Y
2. Y causing X
3. Another factor could be causing both X and Y
   = can only conclude there is a relationship, experimental designs are needed for causality

Question 12

What is the proportionate reduction in error/coefficient of determination?

Answer 12

It is r squared (r) which produces a more meaningful outcome = reveals proportion of variability in DV that is related to the IV. E.g. if r = 0.8, r will = 0.64/64% (proportion of variance accounted for) and ERROR/ unexplained variance = 36% (100=64).