In the regression model belowwe regress a student's
\fQUESTION 5 Spoints SaveAnswer In the regression model below, we regress a student's test score on hisfher tather's education level and hisfher mother's education level. Test Score = 31 + 32 *2 educfathe'r +33 * educ_m.othe'r +14 Based on this regression, how do we access the effect ofparents' education level on child's test score at 10% signicance level? 0 Use a t-test torHo: 32 = 0 againstHB: 32 at 0 ati'li: level. and use a ttesttor Ho: 32 = 0 against Hg: 32 qt 0 at 10% level. Reject if at least one of the ttests rejects. C) Use at-testt'or1'-1':,z = 0 againstHB: 132 i 0 at 5% level, and use a Host forH: 32 = 0 againsl H5: .32 7k 0 at 5% level Reject ifat least one ofthe t-tests rejects C Use a F-test tor the overall signicance of the regressors at 10% level. 0 None at the stated options is coned. QUESTION 6 5points SaveAnswer ln Ihe regression model below, we regress an individual's years of educalion on hither father's years of education and hisiher mother's years of education. education = 31 +32 t educfarher + 33 2 educ_nwthe?" + at What would the restridion 32 + 33 = 1 imply'? C If both parents stayed in school for one more year. we expect their child to complete one more year as well. 0 On average, a child's education equals the sum of both parents' education. 0 If one parent stayed in school for one more year while the education level ofthe other parent was held constant, we expect their child to complete one more year as well. 0 If the mother completed an additional year of education while the father did not, that would have the same predicted effect on the child as iflhe father completed an additional year of education while the mother did not. QUESTION 7 In a multivariate regression, the estimated coefcient b}- is more accurate when 0 the variance of error term is larger, and the regressor x}- ean be very well explained by all the other regressors O the variance of error term is larger, and the regressor x}- ean not be very well explained by all the other regresso rs O the variance of error term is smaller, and the regressor x}- can be very well explained by all the other regressors O the variance of error term is smaller, and the regressor x}- can not be very well explained by all the other regressprs QUESTION 8 Suppose that we estimate the following model Earn =131 +325 +33 exp +34ren +35 urban + or using a sample of 26 observations. And we are interested in testing 140:,3430 Ha: 134:\") at 5% level. What is the critical "value in this test? Write your answer in number and round it to 4- decimal places. QUESTION 5 Suppose we want to estimate the following model 1\" = :31 +3237: +3333 +3434 +151 We will NOT be able to compute The IDLE estimators when ID Co'r'rfY,Xz + 2133} = 1 ID CorriEYm} = 1 C} Corr{32,23 24} = 1 ID CorrXz,u} = 1 QUESTION 1|]I 1li'llhich of the following is true in a regression model that has an intercept term? (3, The Fstetistic for testing the overall signicance of the regressors is always positive. (3, The enor term is always uncorrelated with the regressors. (j The R2 always increases as we add more regressors to the model. (3, None of the stated options is correct
