7. MRA with qualitative information

Question 1

How do you find percentage change with a dummy variable?

Answer 1

You can’t take the log of a dummy variable as it’s binary, instead you use a log-level model

Question 2

How should we write our dummy variables?

Answer 2

For example, in a study of individual wage determination, we might define female to be a binary variable taking on the value one for females and the value zero for males. The name in this case indicates the event with the value one. In this case males would be your base case that you are comparing against

Question 3

What is the base group/ benchmark group?

Answer 3

The group against which comparisons are made

Question 4

How can you use dummy variables for multiple categories?

Answer 4

1) Define membership in each category by a dummy variable
2) Leave out on category which then becomes the base category

Question 5

What is ordinal data?

Answer 5

Ordinal data is a categorical, statistical data type where the variables have natural, ordered categories and the distances between the categories are not known

Question 6

How should we test significance for dummy variables?

Answer 6

The mechanics don’t change, continue to use the t-test

Question 7

What does an interaction variable show in the case of interacting female and marriage when female and marriage are appear separately already?

Answer 7

The additional marriage penalty for female compared to male. It allows the marriage premium to depend on gender.

Question 8

What happens when you interact a dummy variable with a continuous variable?

Answer 8

This results in a change in slope

Question 9

What does an unrestricted model have?

Answer 9

It contains a full set of interactions

Question 10

What does a restricted model have?

Answer 10

The same regression for both groups (including the intercept)

Question 11

What are interaction variables?

Answer 11

Interactions are the two variables multiplied by eachother, for example female and married