4: Non-Linear Regressions

Question 1

interaction terms

Answer 1

allowing for non-zero cross-derivatives of Y with respect to X1 and X2

effect of one explanatory variable on the outcome depends on another explanatory variable

Question 2

categorical variables

Answer 2

categorical variables with more than 2 groups should be broken up into separate dummy variables (number of categories -1)

Question 3

all dummy variables are?

Answer 3

perfectly collinear with no variation

Question 4

linear-log model

Answer 4

1% change in X leads to a change in Y of 0.01β1

Question 5

log-linear model

Answer 5

unit change in X implies (approximately) a constant proportional change in Y equal to β1

change in X of one unit is associated with a 100%β1 change in Y