2. Logistic Regression

Question 1

Why use logistic rather than hierarchical regression?

Answer 1

HR’s don’t work for categorical outcome variables

Question 2

What are the three different forms of LR?

Answer 2

1. Binomial (2 DV groups) 2. Multinomial (various levels of the same factor) 3. Ordinal (2nd v 1st, 3rd v 2nd etc.)

Question 3

When is dummy coding required for an IV

Answer 3

If more than two levels exist

Question 4

What is the benefit of R-squared LL compared with Cox & Snell’s R2 and Nagelkerke’s R2?

Answer 4

It’s not sensitive to number of cases like the other two

Question 5

What it the significance test of choice for individual LR predictors?

Answer 5

Wald’s test

Question 6

What should be taken into account when considering an Odds Ratio?

Answer 6

The prevalence in each group