Lecture 6: Logistic Regression (Alternative)

Question 1

What type of outcome does linear regression model?

Answer 1

Continuous outcome

Question 2

What type of outcome does logistic regression model?

Answer 2

Categorical outcome

Question 3

What assumption does linear regression make about relationships?

Answer 3

Linear relationship

Question 4

Can logistic regression handle non-linear relationships?

Answer 4

Yes

Question 5

Does logistic regression require normally distributed predictors?

Answer 5

No

Question 6

What function does logistic regression use to model probabilities?

Answer 6

Logistic regression uses the logistic (or sigmoid) function to model probabilities.

This transforms the linear combination of predictors into a value between 0 and 1, which can be interpreted as a probability.

• It ensures valid probability predictions.
• It models the log odds of the outcome as a linear function of the predictors.

Question 7

What is the range of values produced by the logistic function?

Answer 7

0 to 1

Question 8

What must outcome categories be in logistic regression?

Answer 8

Mutually exclusive and exhaustive

Question 9

What happens if outcomes are not mutually exclusive or exhaustive?

Answer 9

Model assumptions are violated

Question 10

What are logistic regression coefficients initially expressed as?

Answer 10

Log odds

Question 11

What are odds in logistic regression?

Answer 11

Ratio of event likelihood to non-event likelihood

Question 12

What does exponentiating a coefficient b yield?

Answer 12

Odds ratio (e.g., if exp(B) = 1.5 → each extra date increases odds of being in a relationship by 50%)

Question 13

What does exp(b) = 1.5 mean?

Answer 13

Odds increase by 50% per unit increase in predictor

Question 14

Are changes in probability constant in logistic regression?

Answer 14

No

Question 15

Where is the probability change largest in a logistic curve?

Answer 15

Near 0.5

Question 16

What does b = 0 mean in log odds interpretation?

Answer 16

No effect of predictor

Question 17

What does exp(b) = 1 mean in odds ratio interpretation?

Answer 17

No effect of predictor

Question 18

What is the null hypothesis in logistic regression?

Answer 18

Predictor has no effect on the outcome

Question 19

What test checks if a coefficient significantly differs from 0?

Answer 19

Wald Test

Question 20

What type of distribution is used in the Wald Test?

Answer 20

Chi-square distribution

Question 21

What does it mean if the CI for exp(b) excludes 1?

Answer 21

Predictor is significant

Question 22

What does it mean if the CI for exp(b) includes 1?

Answer 22

Predictor is not significant

Question 23

What is a pseudo-R² used for in logistic regression?

Answer 23

Assessing model fit

Question 24

Name two types of pseudo-R².

Answer 24

Cox & Snell, Nagelkerke

Question 25

What test evaluates model classification accuracy?

Answer 25

Hosmer-Lemeshow test

Question 26

What does a significant Hosmer-Lemeshow test indicate?

Answer 26

Poor model fit

Question 27

What does the omnibus test evaluate?

Answer 27

Whether all predictors improve the model

Question 28

What method adds all predictors at once?

Answer 28

Direct (Enter) method

Question 29

What method adds predictors in blocks?

Answer 29

Sequential (Hierarchical) method

Question 30

What method selects predictors based on statistical criteria?

Answer 30

Stepwise method

Question 31

What does forward stepwise start with?

Answer 31

No predictors

Question 32

What does backward stepwise start with?

Answer 32

All predictors

Question 33

What does each step in stepwise regression involve?

Answer 33

Model comparison

Question 34

What type of outcome does multinomial logistic regression model?

Answer 34

Outcomes with more than two categories

Question 35

How does multinomial logistic regression handle multiple categories?

Answer 35

By using dummy coding