Residual Analysis

Question 1

2 main assumptions with a scatter plot

Answer 1

• Homoscedasticy
• E (u | x) = 0 - whether the model is correctly specified

Question 2

What does homoscedasticity tell us?

Answer 2

• Whether the variance in the model is constant
• Whether the residuals increase/decrease in variance as X increases/decreases

Question 3

What does E (u | x) = 0 tell us?

Answer 3

• Whether or not there are positive/negative systematic deviations from zero

Question 4

How would you tell if a scatter plot violates the E (u | x) = 0 assumption?

Answer 4

• If there is a pattern amongst the data
• could be between any two points in the plot

Question 5

What assumption is explicitly referred to when determining adequacy of a model?

Answer 5

E (u) = 0 - expected value of the error term is equal to zero

Question 6

What assumption is related to the Jarque Bera Test?

Answer 6

The normality assumption

Question 7

What is the normality assumption?

Answer 7

• The residuals of a regression model being normally distributed
• if they are not normally distributed then assumption is violated and null is rejected

Question 8

How can you tell from a Jarque Bera test whether or not the normality assumption has been violated?

Answer 8

• When the p-value is below significance level
• reject the null hypothesis