values of lues of income Table and applying least squares to the first 100 obser
ID: 3159971 • Letter: V
Question
values of lues of income Table and applying least squares to the first 100 observations the second 1o0 observations, yields the sums of squared snd again 258 HETEROSKEDASTICI Sample Includ White b) Ondering the observations according to descending va Use the Goldfeld- Quandt test to test for heteroskedast specifications of the null and alternative hypotheses. those from least squares generalized least squares unc SSES 1.0479 x 107 to test tive hhose and those from for heteroskedastic errors. Include for henotheseleast squares, sset-2.9471x 107 sets of estimates, those from least der the assumption ' 1.2 containsw tion o,..'one, age, and the with White's standard errors, and those (c) Table 11.2 contains three sets of estimates, those d on income, age, and the miles traveled depend on income,a o How do vacation mi (Gi) How do White's standard errors compare with the least standard errors? Do they change your assessentoftheqare of estimation? number of kids in the household?come, lized least squares estimates (ii) Is there evidence to suggest the generalized least In Exercise 7.8 an equation used for the valuation of homes in rounding Boston was estimated. Re-estimating that equation wi standard errors yields the output in Table 11.3. (a) For the coefficients of CRIME, ROOMS, AGE, and TAX. are better estimates? in towns sur- n with White's AGE, and TAX, compare 95% confidence intervals obtained using the standard errors fr compare 11.11 11.1 cise 7.8 with those from Table 11.3. Tablc 11. 2 Estimates for Exercise 11.10: Vacation Model Least squares eatimates VARIABLE T-RATIO STANDARD ERROR BSTIMATED COEFPICIENT 14.201 15-741 -81.826 -391.ss 196 DF AGE KIDS CONSTANT 1.800 3.757 27.13 169,8 7.889 4.189 -3.016 -2.306 Least squares estimates with White standard errors VARIABLE NAME ESTIMATED COEFFICIENT 14.201 15.741 81.826 T-RATIO 196 DF 1.919 3.926 28.86 141.2 7.399 4.010 -2.835 -2.773 AGE KIDS -391.55 Generalized least squares estinates VARIABLE NAME INCOME AGE KIDS ESTINATED COEFFICIENT 13.971 16.348 STANDARD ERROR 1.648 3.422 24.74 145.7 T-RATIO 196 DF 8.476 -78.363 CONSTANT 408.37 -3.168 -2.803Explanation / Answer
C
(i) We say that dependent variable is directly proportional to the independent variable if the coiefficent in estimates is positive. Observing the given table we find that all three sets of estimates suggest that vacation miles travelled is directly related to household income and average age of all adults members but inversely related to the number of kids in the household. As the coieffiecent for number of kids is neagtive.
(ii) The White’s standard errors are very similar in magnitude to those from the least squares standard errors as vissible from the given table. This leads to corresponding similar t-statistics. Thus, the White’s standard errors do not change the assessment of the precision of the estimation. Beasue they can change the precision of estimation only if there is significant difference and different T statastics.
(iii) The generalized least squares estimates and standard errors are also very similar in magnitude to those from least squares. The standard errors are slightly less which may suggest generalized least squares is better because estimate with least standard error will be the best one.
Related Questions
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.