Question

13. Suppose that you run a regression to determine whether gender or race has any significant impact on scores on a test of the economic understanding of children.1 You model the score of the ith student on the test of elementary economics (S;) as a function of the compos- ite score on the Iowa Tests of Basic Skills of the ith student, a dummy variable equal to 1 if the ith student is female (0 otherwise), the aver- age number of years of education of the parents of the ith student, arid a dummy variable equal to 1 if the ith student is nonwhite (0 oth erwise). Unfortunately, a rainstorm floods the computer center and makes it impossible to read the part of the computer output that identifies which variable is which. All you know is that the regression results are (standard errors in parentheses) S.-5.7-0.63%.-0.22x2i + 0.16X3i + 1.20X4i (0.63) (0.88) (0.08) (0.10) N24 R2 .54 a. Attempt to identify which result corresponds to which variable. Be specific. b. Explain the reasoning behind your answer to part a above. c. Assuming that your answer is correct, create and test appropriate hypotheses (at the 5 percent level) and come to conclusions about the effects of gender and race on the test scores of this particular . sample. d. Did you use a one-tailed or two-tailed test in part c above? Why?

Answer 1

sl Gven 31 (0-63) 上 24 an a ton tn Con sesppud to tlhe. eme Js Lommnion Given -Aue Same test Seones, non-w1def kmale students we 'se ve.y?se 已 Vaniab bias, t obviously, Jhe other wiolole X2; atepnese as he. in the Vart eans

heru leu anents o educ ivVoweMent ente Scoate in ependen the ecouoiiCS. ,dentified Sone Lousa fest, xu basic Concepts 6s--he eco omis Suhed-. eest veni おtinated co ebident our statisti

thesis, te computed volbe the nut T to le vhene the nespet O W e Com

ne Coti The null potesis loun pecke expected to be nertive, tuerte,.the He aïable 22· e. Con Otul 0:22 0.25 The

t' value Vaniable isay insi null , altern tnafive HOX2, he uoesfi cient value Stan dare X2. 2.6b ne Mu wel a Sniti cante wsifh an pected ign o ho potue:ce The van!able χǐ k expect 0

き、20 2.oo 1.229 - 12.bo Significou ațthe 5 tev uth o epected si one ot test is used -b test -he Cock^iciet l Pasive*^ofve ive- Two ci en te epen dent Vaialle Ex pec dependent w%uble, one-Hail -test uus used tuted the ANY DOuG HT TLEAE CMENT