lecture 14

Question 1

whats the 1st solution for autocorelation

Answer 1

using non-linear least squares

Question 2

how does non-linear least squares solve the issue

Answer 2

we chose estimates which minimise SSR

Question 3

when do you use the quasi-differencing method

Answer 3

when there are no lagged dependant variables

Question 4

what happen when you omit a variable which is serially correlated

Answer 4

result in a serially correlated error

- so OLS will be biased inefficient and SE of coefficient will be unreiable