Econometrics. a Regression Analysis Essay

2926 Words Oct 2nd, 2012 12 Pages
Question 1:
Run the regression Report your answer in the format of equation 5.8 (Chapter 5, p. 152) in the textbook including and the standard error of the regression (SER). Interpret the estimated slope parameter for LOT. In the interpretation, please note that PRICE is measured in thousands of dollars and LOT is measured in acres.

Model 1: OLS estimates using the 832 observations 1-832
Dependent variable: price

VARIABLE COEFFICIENT STDERROR T STAT P-VALUE

const 119.575 1.54566 77.362 <0.00001 *** lot 1.38850 0.209083 6.641 <0.00001 ***

Mean of dependent variable = 122.076 Standard deviation of dep. var. = 44.3478
…show more content…
Part 1 – Lot coefficient = 1.3885
Part 2 – Lot coefficient = 1.71129

* After analyzing the lot coefficient in both part one and part two, it appears that there is a substantial difference between the two models. The information shows that a one unit increase leads to a price increase of $1338.50 in part one and a price increase of $1711.29 in part two. This information also implies that the model in part one suffers from an important omitted variable bias which can be explained by the additional independent variables in part two. * The two conditions for omitted variable bias are: * The omitted variable or variables must impact the dependent variables, price, based on common sense. In this scenario, the number of bedrooms, bathrooms, fireplaces, etc. would increase the price of the house. * The omitted variable or variables must be correlated with the included independent variable, or variables. In this scenario, as lot size increases, so will all of the other independent variables from part two.

Question 4:
As mentioned in the textbook, OLS standard errors are referred to as homoskedasticity only standard errors. This is because OLS standard errors are strictly valid in the presence of homoskedasticity but they are not valid in the presence of heteroskedasticity. What is heteroscedasticity? What is the effect of heteroskedasticity on the OLS estimates and

Related Documents