EC15 Exam1 Answers s10 PDF

Economics 15
Basic Econometrics
Professor Zabel
Fall 2010
Exam 1 Answer Key
1.(30) Students who apply to college typically need to take the SAT exam.
Exams are given in math and reading and scores on the exams range between
200 and 800. Suppose that you are interested in estimating the effect on hours
spent in an SAT preparation course (HOURS) on total SAT scores (SAT). The
population is all college-bound high school seniors for a particular year.
A.(10) Suppose you are given a grant to run a controlled experiment. Explain
how you would structure the experiment in order to estimate the causal effect
of HOURS on SAT. Explain how you would use the results of the experiment
to estimate this causal effect.
Answer: I would take a random sample of all college-bound high school
seniors. Then I would randomly divide the sample of students into two groups
and each group would receive a different number of hours of an SAT
preparation course, say h1 > h2 (note that h2 could be zero). Because of this
randomized experiment the number of SAT preparation course hours would be
uncorrelated with any other variable, including the error term. In this case the
across the
difference in the average value of SAT scores, diff =
two groups can be attributed to the difference in HOURS. That is, the effect of
studying h2 versus h1 hours would lead to a change in the SAT score of diff.
Consider a more realistic case where students choose how much time to spend
in a preparation course, and you can only randomly sample SAT and HOURS
from the population of all college-bound high school seniors. In this case, the
population model can be expressed as
B.(5) Write down the interpretation of

? Why?
and
. What is the expected sign of
Answer:
is the average value of SAT given no hour spent in the
preparation course.
An increase of one hour in a preparation course will lead to a change in SAT
score of , on average.
The expected sign of is positive. This is because an additional hour of
preparation should provide the person with greater knowledge to better on the
SAT test.
C.(5) Explain how you would determine if this effect is
statistically/economically significant
Answer: To determine if this effect is statistically significant, test the
hypothesis
A rejection of this hypothesis will indicate that the impact is statistically

significant.
To determine if this effect is economically significant, calculate the
standardized coefficient
I would NOT calculate the elasticity since the units of measurement of SAT is
not meaningful.
D.(5) Do you think that the above model allows one to measure the causal
relationship between HOURS and SAT? Why or why not? Be as specific as
possible in giving reasons for your answer.
Answer: It is unlikely that measures the causal relationship between
HOURS and SAT. This is because the students that take the preparation
course are not a random sample from the population of students who will take
the SAT. For example, one might expect that these students have a lower GPA
on average than students who do not take the course. Thus, one could get the
result that the average SAT score for the students who take the course is lower
than for the students who do not take the course and hence the estimate of
could be negative.
2.(10). What is the standard error of the regression? (Explain in words, just
writing down the formula is not good enough). How does it relate to the
goodness of fit of the regression? Explain how the SER is used in hypothesis
testing.
Answer: The standard error of the regression (SER) is an estimator of the
standard deviation of the error term in the linear regression model. The
formula is
Since the variance of y conditional on x is equal to the variance of the error

term, the standard error is an estimator of how spread out the observed values
of Y are around its conditional (on X) mean conditional. So the larger the
SER, the worse the fit of the regression. When carrying out hypothesis tests
about the coefficients in the regression model using the OLS estimators of
these coefficients, we need the standard deviations of the OLS estimators.
These standard deviations are a function of the standard deviation of the error
term. Since the standard deviation of the error term is not observed, we use the
SER to estimate it.
3.(50). Suppose you work for the Justice Department and you are investigating
price discrimination in the retail market. In particular, you are interested in
determining if fast-food restaurants charge higher prices in the areas with
larger concentrations of blacks. You collect data at the zip code level on prices
for various items at fast food restaurant along with characteristics of the
population in each zip code in New Jersey and Pennsylvania. You form the
following model
A.(5) Interpret the coefficient estimates for pctncar, income, and hrsopen.
Answer: A one percentage point increase in pctncar will lead to a $0.0059
decrease in the price of an entre, on average and all other regressors held
constant.
A $10,000 increase in median family income will lead to a $0.063 decrease in
the price of an entre, on average and all other regressors held constant.
If stores stay open one more hour per day, the price of an entre will decrease
by $0.143, on average and all other regressors held constant.
B(10). Show that the R2 for the regression is 0.412 and interpret this result.
Carry out a test for the overall significance of the regression at the 1%
significance level. Interpret the result of this test.
Answer:
This implies that 41.2% of the variation in the price if an entre is explained by
the regression.
The test for the overall significance of the regression at the 1% significance
level
The result is
Hence, we reject the null hypothesis that the slope coefficients are jointly zero.
This means that the regressors significantly explain the variation in pentree.
C(10). Test the hypothesis at the 1% significance level that there is price
discrimination against blacks in fast food stores in New Jersey and
Pennsylvania. What is the result? Is the result economically significant (why
or why not)?
Answer:
The result is
The result is that there is evidence of discrimination against blacks in fast food
stores in New Jersey and Pennsylvania.
To determine economic significance, we calculate the elasticity
Or the standardized coefficient
Thus, using the elasticity, the impact of pctblack on the price of an entre IS
NOT economically significant. BUT using the standardized coefficient, the
impact of pctblack on the price of an entre IS economically significant
D.(10). Calculate the expected difference in the price of an entree for zip codes
that differ in the median household income by $20,000. Test at the 5%
significance level the hypothesis that the magnitude of this difference is at least
10 cents. Calculate the p-value of the test and interpret your result.
Answer:
E[entre|income = x + 2] - E[entre|income = x] = 2
=2 -0.063=-0.126
Hypothesis test:
The result is
The p-value = P(Z > 0.44) = 0.33. This implies that the smallest significance
levels at which the null hypothesis will be rejected is 0.33.
E.(10). The variables pctpov, pctncar, and crimerate capture local economic
conditions. Carry out a test at the 5% significance level of whether or not the
local economic conditions affect the price of an entre.
Answer:
The result is
Variables
pctblack
income
crimerate
wagest
hrsopen
pctpov
pctncar
nstores
Constant
Table 3: Regression Results

Dependent Variable: Price of an Entree
(1)
0.0070
(0.0020)
0.0630
(0.0296)
1.4208
(0.8869)
0.0111
(0.0798)
0.1435
(0.0098)
0.0120
(0.0120)
0.0059
(0.0064)
0.0651
(0.0163)
3.8355
(0.4344)
377
Observations
R squared
Standard error of the Regression
Total Sum of Squares
Standard errors in parentheses
(2)
0.0050
(0.0016)
0.0135
(0.0230)
0.0285
(0.0796)
0.1435
(0.0097)
0.0513
(0.0148)
3.4235
(0.3949)
377
0.400
0.5018
157.61

EC15 Exam1 Answers s10 PDF

Uploaded by

Copyright:

Available Formats

EC15 Exam1 Answers s10 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EC15 Exam1 Answers s10 PDF

Uploaded by

Copyright:

Available Formats

What factors are considered in estimating the causal effect of SAT preparation hours on SAT scores?

What is the interpretation of the coefficients in the population model for estimating the effect of SAT preparation hours on SAT scores?

Economics 15

B.(5) Write down the interpretation of

. What is the expected sign of

A rejection of this hypothesis will indicate that the impact is statistically

Since the variance of y conditional on x is equal to the variance of the error

Or the standardized coefficient

Table 3: Regression Results

You might also like