Question

Remaining Time: 2 hours, 46 minutes, 47 seconds. Question Completion Status: QUESTION 25 Use the following information to answer questions 25-28: A cross-section model for expenditures for police protection per capita in major cities in the U.S. is estimated for the year 2005 using 45 observations, one for each city. The model takes the form: Y = B1 + B 2 X 2i+ B3X 3i+ & Where: Yi - per capital expenditures on police protection in city i X 2i = income per capita in city i X 31 = population density of city i. The model is estimated with OLS. However, the estimated model is thought to suffer from heteroscedasticity. To evaluate this (the presence of heteroscedasticity) the data are rearranged by the size of X 3, the variable thought to be correlated with the variance of the error term. The middle 1/5 of the data are eliminated, i.e., d=(1/5)n. Two regressions are then estimated, one with the low values of X3 and one with the large values of X3. The estimated equations are: For low values of X3: Y = 32000 + 214 X 21 + 3500 X 3; and ESS 1 = 21.9 For large values of X3: Y = 214 + 319 X 2i + 1725 X 3; and ESS 2 = 42.5 To test for heteroscedasticity, we use Generalized-Difference Test Goldfeld-Quandt Test Durbin-Watson Test First-Difference Test Click Save and Submit to save and submit. Click Save All Answers to save all answers. Save All Answers Co DOLL

please help with this econometrics question

Answer 1

Q25) The answer is (b) Goldfred-Quandt test as it is used to test for the presence of heteroskedasticity in the residuals.

Rest all options are wrong as they do not test for heteroskedasticity