L5 - Regression Analysis II

Question 1

What is a spurious correlation?

Answer 1

Relationship between income and height, but gender is the
responsible factor

Question 2

What is a masked relationship?

Answer 2

Example: Positive relationship between happiness and debt, but
relationship reverses when controlling for income (both happiness and debt are positively related to income)

Question 3

How can spurious correlations and masked relationships be detected?

Answer 3

By doing multiple linear regression

Question 4

partial correlation

Answer 4

Question 5

interpretation of the ß

Answer 5

Interpretation of the ß = standardized regression coefficient: A one unit increase in X so a 1 SD increase in X leads to an e.g. .546 units SD in Y.

Question 6

F-test

Answer 6

The F-test of overall significance indicates whether your linear regression model provides a better fit to the data than a model that contains no independent variables.

Question 7

How can you evaluate if the value of the regression coefficients b are significantly different from 0?

Answer 7

t-test

Question 8

How much variance in the outcome variable
is accounted for by the predictor? Which test?

Answer 8

R-squared

Question 9

What can you do if there is a nonlinear relationship between DV and IV?

Answer 9

add a quadratic component.

Question 10

What is a lack of multicollinearity?

Answer 10

Overlap among predictors should not be too large (i.e., there should be no perfect linear relationship between two or more predictors)

Question 11

What happens under multicollinearity?

Answer 11

Under multicollinearity, the regression coefficients may be unstable (i.e., it will be difficult to assess the individual importance of a predictor) and their standard errors large

Question 12

What is a problematic pearson correlatino for multicollinearity?

Answer 12

r > .8

Question 13

What is the tolerance in multicollinearity?

Answer 13

Question 14

What tolerance is a serious problem?

Answer 14

A small tolerance is problematic like .1 (.02 is a potential problem)

Question 15

What is the VIF

Answer 15

Variance Inflation Factor = 1/ Tolerance

• largest VIF should not be greater than 10
• average should not be much greater than 1

Question 16

What can you do if there is an indication for multicollinearity?

Answer 16

Consider dropping redundant predictors.

Question 17

In MLR how many observations per predictor?

Answer 17

min 10

Question 18

What should be observations N exceed the # of predictors by

Answer 18

at least 50

Question 19

Does the recommended sample size depend on purpose of analysis?

Answer 19

Yes.

• if goal is to test model overall: 50 + 8* # predictors
• if goal is to test individual predictors: 104 + # predictors

Question 20

How to determine correct sample size

Answer 20

power analysis in the program

Question 21

What is the forward regression method in exploratory analysis?

Answer 21

Stat with predictor that correlates most strongly with outcome variable, then add predictors that yield largest improvement

Question 22

Are there standardized coefficients ß for dummy-coded predictors?

Answer 22

no. Only for continous predictors

Question 23

3 conditions in mediation analysis

Answer 23

1) Is there a significant relationship between independent variable and mediator (a)?
2) Is there a significant relationship between mediator and dependent variable (b)?
3) Is there a significant relationship between independent and dependent variable (c)?

Question 24

Key analysis in mediation analysis

Answer 24

Key analysis: Is regression weight c reduced when mediator and independent variable are used simultaneously to predict the dependent variable (c*)?

Question 25

What is the question in moderation analysis?

Answer 25

Does relationship between independent and dependent variable differ for different levels
of the moderator?

Question 26

How to test in moderation analysis

Answer 26

Tested by including the independent variable, the moderator, and their interaction as predictors

Question 27

What is important in moderation analysis

Answer 27

IMPORTANT: independent variable and moderator need to be centered (to avoid multicollinearity)

Question 28

Centering image

Answer 28