Lecture 7

Question 1

De Regression equation in general form:

Question 2

What use is a regression with many variables?

Answer 2

1.Testing of theory
2. Investigating how the effect of independent variable X on dependent
variable y differs with and without controlling for other variables
(elaboration logic, see lecture 4)
3. ‘Increasing’explained variance (Dangerous!)
4. Including many dummy variables for categorical independent variables
with many categories
5. Investigating non-linear (curvilinear) relationships
6. Including product variables to investigate interaction/moderation

Question 3

Strategies for constructing your Regression Model

Question 4

Partition of variance in a standard regression

Answer 4

Standard regression:
all variables inserted in the model at once

Shaded areas are the unique contributions to explained variance by X1,
X2 and X3 (squared part correlations)

Question 5

Partition of variance in a Sequential Regression

Answer 5

Sequential Regression
Variable(s) are inserted in the regression model in steps

Shaded areas are the unique contributions to explained variance by X1,
X2 and X3

Question 6

Significance test Overall Regression Model

Question 7

F-test for Model comparison