Week 10 Lecture 10 - hierarchal regression

Question 1

What is zero order ^2?

Answer 1

variance explained by xi, expressed as a proportion of total variance in y

Question 2

What is part ^2?

Answer 2

unique variance explained by xi, expressed as a proportion of total variance in y

Question 3

What is partial^2?

Answer 3

unique variance explained by xi, expressed as a proportion of the total variance in y that remains after the variance explained by the other predictors has been removed

Question 4

How to calculate shared variance?

Answer 4

if just 2 predictors:
zero order correlation ^2 - part correlation ^2

Question 5

How to calculate the variance explained by all predictors combined?

Answer 5

sum(part correlations^2) + shared variance

Question 6

What is this equal to?

sum part correlations^2 + shared variance

Answer 6

r^2

Question 7

How to calculate unexplained variance?

Answer 7

1-r^2

Question 8

How to calculate partial correlation ^2?

Answer 8

unique variance / (unexplained variance + unique variance)

Question 9

How to calculate total variance?

Answer 9

sum part correlations^2 + shared variance

Question 10

What is hierarchal regression?

Answer 10

• predictor variables entered in a specific order of “steps” based on theoretical grounds
• relative contribution of each “step” (set of predictor variables) can be evaluated in terms of what it adds to the predictions of the outcome variable

Question 11

Why use a hierarchal regression?

Answer 11

• to examine the influence of predictor variables on an outcome variable after “controlling for” (partialling out) the influence of other variables

Question 12

What do change statistics tell us?

Answer 12

tell us about the explanatory power of a predictor variable after other predictor have been controlled for (partialled out)