Correlational Designs + Analysis

Question 1

What do correlational designs aim to measure?

Answer 1

Relationship between two variables

Question 2

When should an Enter Method Multiple Regression be used to analyse data?

Answer 2

1. Aiming to predict one variable from scores on other variables
2. Not looking for how a predictor contributes on top of other predictors

Question 3

When should an Heirarchal Multiple Regression be used to analyse data?

Answer 3

1. Aiming to predict one variable from scores on other variables
2. Controlling for a variable

Question 4

What does R squared tell us about when using regression?

Answer 4

How well our combined predictor variables are able to predict the outcome variable

R2 = 1 [Perfect prediction]
R2 = 0 [Terrible prediction]
R2 is usually somewhere inbetween

Question 5

What does R Square Change tell us?

Answer 5

To what extent the addition of the new variables increase R Squared
Conducted after heirarchal regression

Question 6

What does simple linear regression allow us to do?

Answer 6

Predict a score on one variable from a known source on a predictor

Y = a + bx

Question 7

What are standardised beta coefficients?

Answer 7

They tell us about the strength of relationship between predictor and DV

Question 8

What are the six assumptions of multiple regression?

Answer 8

1. Continous outcome variable
2. Linear Relationship
3. Varience presense
4. Multicollinearity (too high correlation)
5. Outliers removed
6. Sample size is large