Question

3. Consider you are interested in estimating the determinants of earnings. For this purpose you estimate a regression of annual earnings (EARN , in $000's) as a function of the years of education EDUC, with 12 being high school, 16 a bachelor degree, etc.) and the years of experience EXP number of years working). The estimated regression you obtain from a sample of one thousand individuals is (standard errors in parenthesis): EARN;= -50.3 +3.0*EDUC +1.5*EXP (0.42) (0.44) 7.1 3.4 RP=0.45 N=1,000 a What is the expected sign of the coefficients? Education: Ans: + (positive) Experience: Ans: + (positive) b. What is the interpretation of the R?? c. What is the interpretation on the coefficient of education? 2 d Which provides higher earnings, one more year of education or one more year of experience? e Use a 5% significance level to test the following hypothesis. Show all steps. Ho: BEXP < 0 HA: BExp > 0 f. Calculate a 95% confidence interval for the coefficient of education. (Hint: Remember, confidence intervals use two-sided critical values.)

Answer 1

(a from the given Information, + positive 6 R² = 0.45 45% of the Variation in annual earning is explained by the independent variables Education & expenance. © Slope (B) r Education 3 (d) There is rise in lunit of education then we predicted annual earning is increases by 3 unit when experiance is constant Experiance provide higher Earnings © To test the hypothesis, Hoo Bearso Vs Hi: Bedc > 0 Test statistics, t= ßi SECÊ, ) 3 0.42 t=7.14 P-value- Plt >7.14] lesing t-table, Pvalue= 0.0000

Decision : Here, P-value <=0.05 we Reject Ho at 5% level. Conclusion : There is sufficient exidance to Support the claim that coefficient of education is positive (Bi>O) (6) The 95% contidance intewal for B, t ta in 2 s SECẾ ) ) [3 I to.025, 993 (0:42) 키 (1.965) (0-42)] sty, using t-table → [ 3 I 0.8253) [ 2.18 3.83)