Question

A researcher is examining the effect of number of years in a particular job (tenure) on the hourly wage (USD) earned. She estimates two regression models using data on 124 young women surveyed in 1988. In the first regression, she regressed logged hourly wage on logged tenure in the current job (years in current job). In the second regression she regressed logged hourly wage on logged tenure in the current job, as well as age, total years of schooling completed, and total years of work experience in any job). reg In_wage 1n_tenure Source SS df MS = Model 4.96082255 Residual | 32.5291117 1 4.96082255 122 .266632063 Number of obs F(1, 122) Prob > F R-squared Adj R-squared Root MSE 124 18.61 0.0000 0.1323 0.1252 .51636 11 Total 37.4899342 123.304796213 In_wage 1 Coef. Std. Err. p>It! [95% Conf. Interval] 1n_tenure _cons .163222 1.641472 .0378406 .0704397 4.31 23.30 0.000 0.000 .0883128 1.50203 .2381313 1.780915 · reg In_wage In_tenure age yrs_school experience Source SS df MS 11 4 Model 9.7457874 Residual 27.7441468 2.43644685 .233144091 Number of obs F(4, 119) Prob > F R-squared Adj R-squared Root MSE 124 10.45 0.0000 0.2600 0.2351 .48285 119 Total | 37.4899342 123 .304796213 1n_wage 1 Coef. Std. Err. t P>It [95% Conf. Interval] 1n_tenure age yrs_school experience cons .0959318 .0081925 .0542063 .0291484 .0418616 .0147956 .0154112 .0121425 .6275823 2.29 0.55 3.52 2.40 0.53 0.024 0.581 0.001 0.018 0.600 .0130417 -.0211043 .0236905 .0051049 -.912602 .1788219 .0374894 .0847221 .0531918 1.572749 .3300737

Answer 1

a) The interpretation of the coefficient estimate of 0.0959318 in the second regreassion is that if 1 unit change in log tenure represents 0.0959318 unit increase in the hourly wage.

b) The coefficient on log job tenure is much lower in the second regression compared to the first one since for the first regression we use only one variables that is log tenure to predict log wage. While for the second regression we use additional three variables that is age, years in school and experience along with log tenure to predict lo wage.

c) The test statistic of 0.55 and the associated value of 0.581 coreresponds to the age variable in the second regression.

The formula that we use is T = coefficient value/standard error of the coefficient.

Here, we use coefficent = 0.0081925 and standard error = 0.0147956

The hypothesis used that go with the p-value are-

H₀: The variable is not significant. vs H₁: The variable is significant.