Scribd
Upload
9 views

HW CH 3

Uploaded by

Ahmad Firdaus
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
9 views

HW CH 3

Uploaded by

Ahmad Firdaus
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Econometric Homeworks Chapter 3

Problem 3

(i)

With assump on of tradeoff between sleep and work, the expected sign of 1 should be
minus. It means that when there is an increase in me of work will cause reduc on in
sleeping me and vice versa.

(ii)

Age and Educa on might have a complex way to influence the sleep me of a person. For
example, one might argue that the sleep me decreases in line with the increase of a person
age. However, it might be also influenced by that person ac vity during those age and the
sleep me might also increase in certain level of age (e.g. when a person already reach
pension age). This also true for educa on in a way that a person might have less sleep me
during their advanced educa on. Nevertheless, if a person already finish their educa on
and not doing any other ac vity then the sleep me might not decrease at all. In conclusion,
the sign of 2 and may not be obvious due to those complex rela onship.

(iii)

Five hour per week will be equal to 300 minutes per week. With coefficient of 1=-0.148,
then the predicted value of sleep will fall/decrease by 44.4 minutes per week (-0.148 x 300
minutes). For a week period, a decrease less than an hour might not overwhelming.

(iv)

The sign means that an increasing educa on will decrease the sleeping me. For example if
there is an increase in educa on by 1 year then the sleeping me will fall by 11.13 minutes
per week. In span of 1 year, a decrease of 11 minutes is not a significant amount.

(v)

With R-Square of 11.3%, the three variables are not explain much about sleep. Another
factors that may influence sleep are marriage status, health condi on, and children. These
other factors may correlated to totwork. For example, having more children might influence
people to work more in order to increase income.
Problem 4

(i)

Higher number in rank means lower performing students. Thus, lower performance
students are expected to have lower salary and cause 5 expected value to be nega ve.

(ii)

a. For LSAT and GPA, higher score means be er performing students. Thus, lower
performance students are expected to have higher salary and cause 1 and 2 expected
value to be posi ve.

b. For libvol and cost, the rela on of those two variables with salary might be unclear. The
reason is these two variables do not directly influence hiring company to decide salary of
the graduates. Moreover, it more likely that libvol and cost will influence the performance
of the students rather than directly to the salary.

(iii)

An increase of one point in GPA will cause an increase salary by 24.8 percentage point.

(iv)

Coefficient of libvol has posi ve sign which means that an increase in number of volumes in
library associated with an increase in salary.

(v)

Yes, it will be be er to enter higher ranked law school (higher ranked means lower number)
because according to the regression result it will increase the salary. If there is a difference
of 20 number, the salary will differ by 6.6 percentage point.
Problem 7

The cause of OLS es mators to be biased will be (ii) omi ng an important variable. It will
cause the omi ed variable bias which cause the model to not sufficiently explain how the
independent variable affect the dependent variable.
Computer Problem 5

First par alling out step: regress educ on exper and tenure

The result is

educ = 13.57 -.074 exper + .048 tenure + r1

n = 526, R2 = .101

Second step: regress wage on residual

The result is

Wage = 5.896 + 0.599 r1


This coefficient of r1 is similar to educ coefficient if we regress wage on educ, exper, and
tenure

You might also like