Question

1. [30 marks] A student of ICEF considers for her graduation thesis the hedonic price function for houses. The hedonic price refers to the implicit price of a house given certain attributes (e.g., the number of bedrooms). The data contains the sale price of 546 houses sold in the summer of 2017 in the Moscow region (in thousands of rubles) along with their important features. The following characteristics are available: the lot size of the property in square meters (lotsize), the numbers of bedrooms (bedrooms), the number of full bathrooms (bathrooms), and a dummy indicating the presence of air-conditioning (airco). Consider the following ordinary least squares results log(price)?i = 9.894 + 0.400 log(lotsize)i + 0.078 bedroomsi + 0.096 bathroomsi + 0.212 aircoi n = 546 (1) (0.232) (0.028) (0.015) (0.023) (0.024) [0.233] [0.028] [0.017] [0.024] [0.023] The usual standard errors are in parentheses, the heteroskedasticity robust standard errors are in square brackets, and RSS measures the residual sum of squares. 1.1. [10 marks] ? Interpret the parameter estimates on log(lotsize), bedrooms and airco. Discuss also the intercept. Comment the differences in standard errors of two types. ? The variable airco is a dummy variable so it can take only two values so interpretation based on the properties of infinitely small values above is not quite correct. Calculate the effect of having airconditioning on the price of a flat and compare it with the result e based on the analysis of differentials obtained above. 1.2. [10 marks] ? The student suspects the presence of heteroscedasticity in the data. She noticed that the variance of log(price) is increasing with log(lotsize). To test for heteroscedasticity she orders the 546 observations by the magnitude. Then she runs the same regression (1) using 212 observations with the smallest values of log(lotsize), obtaining RSS1 = 7.781, and then runs the same regression for 90 observations with the largest values of log(lotsize), obtaining RSS2 = 6.214 (an unequal number of observations was taken due to the fact that number of lots with high prices turned out to be significantly less). Help the student to run the appropriate test and to make the conclusion. 1.3. [10 marks] ? Student's friend advised her to use White test instead. She runs test for restricted regression (2) obtaining for the auxiliary equation of the White test (with cross terms) the value of R2 = 0.0317. Help the student to complete the test briefly explaining her the logic of this test, indicating distribution, degrees of freedom, critical values and final conclusion.

          1. [30 marks] A student of ICEF considers for her graduation thesis the hedonic price function for houses. The hedonic price refers to the implicit price of a house given certain attributes (e.g., the number of bedrooms). The data contains the sale price of 546 houses sold in the summer of 2017 in the Moscow region (in thousands of rubles) along with their important features. The following characteristics are available: the lot size of the property in square meters (lotsize), the numbers of bedrooms (bedrooms), the number of full bathrooms (bathrooms), and a dummy indicating the presence of air-conditioning (airco). Consider the following ordinary least squares results

log(price)?i = 9.894 + 0.400 log(lotsize)i + 0.078 bedroomsi + 0.096 bathroomsi + 0.212 aircoi n = 546 (1)
(0.232) (0.028) (0.015) (0.023) (0.024)
[0.233] [0.028] [0.017] [0.024] [0.023]

The usual standard errors are in parentheses, the heteroskedasticity robust standard errors are in square brackets, and RSS measures the residual sum of squares.
1.1. [10 marks] ? Interpret the parameter estimates on log(lotsize), bedrooms and airco. Discuss also the intercept. Comment the differences in standard errors of two types.
? The variable airco is a dummy variable so it can take only two values so interpretation based on the properties of infinitely small values above is not quite correct. Calculate the effect of having airconditioning on the price of a flat and compare it with the result e based on the analysis of differentials obtained above.
1.2. [10 marks] ? The student suspects the presence of heteroscedasticity in the data. She noticed that the variance of log(price) is increasing with log(lotsize). To test for heteroscedasticity she orders the 546 observations by the magnitude. Then she runs the same regression (1) using 212 observations with the smallest values of log(lotsize), obtaining RSS1 = 7.781, and then runs the same regression for 90 observations with the largest values of log(lotsize), obtaining RSS2 = 6.214 (an unequal number of observations was taken due to the fact that number of lots with high prices turned out to be significantly less). Help the student to run the appropriate test and to make the conclusion.
1.3. [10 marks] ? Student's friend advised her to use White test instead. She runs test for restricted regression (2) obtaining for the auxiliary equation of the White test (with cross terms) the value of R2 = 0.0317. Help the student to complete the test briefly explaining her the logic of this test, indicating distribution, degrees of freedom, critical values and final conclusion.

1. [30 marks] A student of ICEF considers for her graduation thesis the hedonic price function for houses. The hedonic price refers to the implicit price of a house given certain attributes (e.g., the number of bedrooms). The data contains the sale price of 546 houses sold in the summer of 2017 in the Moscow region (in thousands of rubles) along with their important features. The following characteristics are available: the lot size of the property in square meters (lotsize), the numbers of bedrooms (bedrooms), the number of full bathrooms (bathrooms), and a dummy indicating the presence of air-conditioning (airco). Consider the following ordinary least squares results

log(price)?i = 9.894 + 0.400 log(lotsize)i + 0.078 bedroomsi + 0.096 bathroomsi + 0.212 aircoi n = 546 (1)
(0.232) (0.028) (0.015) (0.023) (0.024)
[0.233] [0.028] [0.017] [0.024] [0.023]

The usual standard errors are in parentheses, the heteroskedasticity robust standard errors are in square brackets, and RSS measures the residual sum of squares.
1.1. [10 marks] ? Interpret the parameter estimates on log(lotsize), bedrooms and airco. Discuss also the intercept. Comment the differences in standard errors of two types.
? The variable airco is a dummy variable so it can take only two values so interpretation based on the properties of infinitely small values above is not quite correct. Calculate the effect of having airconditioning on the price of a flat and compare it with the result e based on the analysis of differentials obtained above.
1.2. [10 marks] ? The student suspects the presence of heteroscedasticity in the data. She noticed that the variance of log(price) is increasing with log(lotsize). To test for heteroscedasticity she orders the 546 observations by the magnitude. Then she runs the same regression (1) using 212 observations with the smallest values of log(lotsize), obtaining RSS1 = 7.781, and then runs the same regression for 90 observations with the largest values of log(lotsize), obtaining RSS2 = 6.214 (an unequal number of observations was taken due to the fact that number of lots with high prices turned out to be significantly less). Help the student to run the appropriate test and to make the conclusion.
1.3. [10 marks] ? Student's friend advised her to use White test instead. She runs test for restricted regression (2) obtaining for the auxiliary equation of the White test (with cross terms) the value of R2 = 0.0317. Help the student to complete the test briefly explaining her the logic of this test, indicating distribution, degrees of freedom, critical values and final conclusion.

Added by Lauren K.

Question

Please give Ace some feedback