Multiple Linear Regression (3)

Question 1

What else can the independent variables be called?

Answer 1

Regressors or covariates

Question 2

What is the unsystematic part of a regression model?

Answer 2

unsystematic part part u, contains
the part of y not explained by the regressors x1, . . . , xk

Question 3

What is an example of a multiple linear model with non-constant marginal effects?

Answer 3

When the model includes one regressor (x) and also its square value (x^2)

Question 4

What is x*?

Answer 4

There is a value of x so that the marginal
effect is zero, given by
x*=-B1/2B2
only applicable to non-constant marginal effects

Question 5

When will x* minimise/maximise the population regression function?

Answer 5

minimise when B2>0
maximise when B2<0

Question 6

What are the equations needed to find the slope coefficients in multiple linear regression models?

Answer 6

check ppt 2

Question 7

What is the equation to find the intercept given the slope estimates?

Answer 7

check ppt 2

Question 8

What is the equation for the sum of squared residuals in multiple linear regression models?

Answer 8

check ppt 3

Question 9

What is the equation linking the sum of squares, explained sum of squares and sum of squared residuals?

Answer 9

SST=SSE+SSR

Question 10

What is the equation for SST?

Answer 10

check ppt 4

Question 11

What is the equation for SSE?

Answer 11

check ppt 4

Question 12

What does R^2 show and what do different values of it mean?

Answer 12

R^2 is the ratio of the explained variation over the total variation. Thus, it is interpreted as the proportion of the
sample variation in yi that is explained by
y^i
R^2 = 1 means perfect fit

Question 13

What is heteroskedasticity?

Answer 13

If the conditional variance Var(u|x1, . . . , xk ) depends on any of the regressors (not constant)