1. (20 points) In the R output below, log. income is workers log income; be is a dummy variable of blue collar worker type; education is a continues variable of workers years of education.
a. Consider the following regression of log. income on be and education:
model2 <- lm(log.income be + education)
summary(model2)
Call:
lm(formula * log.income be + education)
Coefficients :
Estimate Std. Error t value Pr(>ItI)
(Intercept) 3.202348 0.412498 7.763 1.2e-09 ***
be -0.530492 0.308001 -1.722 0.0924 .
education 0.010466 0.005221 2.004 0.0515 .
Signif. codes: 0 W 0.001 ‘**’ 0.01 V 0.05 0.1 ‘ ‘ 1
From the regression results of model2 above, find the percentage change in income for blue collar
worker type or not. Note: be is a dummy variable.
a. From the regression results of model2 above, find the percentage change in income for one addi-
tional year of education. Note: education is a continues variable.
b. Further consider the following regression of log. income on be. education and be * education:
model3 <- lm(log.income be + education + be * education)
summary(model3)
Call:
lm(formula * log.income be + education + be * education)
Coefficients :
Estimate
(Intercept) 3.589474
be -1.666881
education 0.005398
be¡education 0.034645
Signif. codes: 0 ***' 0.001 '**' 0.01 V 0.05 V 0.1 ' ' 1
From the regression results of model3 above, derive the relationship between log.income and education for non-blue collar workers, i.e., bc=0. In particular, for non-blue collar workers, how large is the percentage change in income for one additional year of education?
a. From the regression results of model3 above, derive the relationship between log.income and education for blue collar workers, i.e.. bc«l. In particular, for blue collar workers, how large is the percentage change' in income for one additional year of education?
Std. Error t value Pr(>ItI)
0.419223 8.562 1.14e-10 ***
0.541789 -3.077 0.00372 **
0.005332 1.012 0.31735
0.013940 2.485 0.01712 *
1