Take Test: ECON2300 Quiz 3 (Semester 1, 2024)
Test Information
Description Due date: 4pm, Friday, March 22, 2024
Please read all instructions carefully before opening this Quiz.
Instructions Please pay close attention to the number of decimal places required (if any) for each answer. The required number of decimal places may differ from question to question.
Avoid rounding during intermediate calculations where possible.
This Quiz is an R-Exercise and it is not timed. This means that you can open the R-Exercise and return to it as many times as you need to (provided that you do not click submit).
There is only one attempt for this R-Exercise.
The R-Exercise is marked out of 7, If this quiz is amongst the best 7 out of the 10 quizzes this semester, it will contribute (30/7)% (approximately 4.3%) towards your final grade.
The closing time for this R-Exercise is 4pm on Friday, March 22, 2024. Please make sure that you have submitted your answers by this time. Remember that you must click submit before the
deadline for your R-Exercise to be marked.
Please Note: If you encounter any technical issues with the R-Exercise, please email cml.2300@uq.edu.au. Do not email Quiz issues to the Course Coordinator or Economics Administration.
Otherwise there may be a delay in responding to your enquiry.
=====================================================
FURTHER INFORMATION ABOUT QUIZ 3:
Data: We will use the dataset "MASchools" included in the package "AER". The dataset contains data on test performance, school characteristics and student demographic backgrounds
for school districts in Massachusetts. (The lectures use a similar dataset for schools in CA.) For the questions in this exercise, therefore, you should load the dataset "MASchools" via:
> data("MASchools", package = "AER")
There are alternative ways to load the data in R. In tutorial sessions for example, we loaded the package first and then loaded the data.
R-tips: You might want to attach the dataset via:
> attach(MASchools)
You do not need to attach the dataset for this assessment. By attaching the dataset, however, you do not have to specify to which dataset the variables belong when you use them. For
example, suppose you want to get summary statistics of the variable salary. Then, before attaching the dataset, you have to do this via
> summary(MASchools$salary)
But, once you attach the dataset "MASchools", you can do the same thing simply by
> summary(salary)
The R command attach() is useful when you work with many variables. When you do not need the dataset anymore, you can detach it via:
> detach(MASchools)
Multiple
Attempts
Not allowed. This Test can only be taken once.
Force
Completion
This Test can be saved and resumed later.
Your answers are saved automatically.
To learn the effect of student-teacher ratio on test scores, we use two variables: (1) student-teacher ratio (STR) and (2) 4th grade score (TestScore),
which is the sum of math, English, and science in the dataset MASchool.
What is the sample mean of the 4th grade score (one decimal place)?
QUESTION 1 1 points   Save Answer
Estimate the population regression model:
TestScore i= β 0+ β 1STR i+ u i
where E[u i |STR i]= 0. What is the OLS estimate of the slope coefficient (two decimal places)?
QUESTION 2 1 points   Save Answer
Consider the estimated regression equation in Question 2. What is the robust standard error for the OLS estimate of the slope coefficient (two decimal
places)? For this question, the standard error type in R has to be either "stata" or "HC1".
QUESTION 3 1 points   Save Answer
The minimum and maximum of STR are 11.40 and 27.00, respectively, in the data. Using the estimated regression equation in Questions 2 and 3, predict
the difference in TestScores between the school district with STR = 11.40 and the school district with STR = 27.00. Report the difference in absolute
value (two decimal places).
QUESTION 4 1 points   Save Answer
Question Completion Status:
Click Save and Submit to save and submit. Click Save All Answers to save all answers. Save All Answers Save and Subm
Note that this exercise replicates Section 5.3 in Stock and Watson, but with the MA data instead of CA data. Hence, it would be helpful to read the
section once again. Generate a new variable D i in R that equals either 0 or 1, depending on whether the student-teacher ratio is less than 17, i.e.,
D i =
⎧
⎪
⎪
⎪
⎪
⎪
⎨
⎪
⎪
⎪
⎪
⎪
⎩
1 if the student− teacher ratio in i thdistrict < 17
0 if the student− teacher ratio in i thdistrict ≥ 17
Estimate the population regression model
TestScore i=α 0+ α 1D i+ e i
where E[e i |D i]= 0. Note that different notations (a 0, a 1, e i ) are used instead of (β 0, β 1, u i ) because the regression model in this question is
different from the one in Question 2. What is the OLS estimate for the conditional expectation,
E[TestScore | STR ≥ 17]
Report your answer to two decimal places.
QUESTION 5 1 points   Save Answer
Using the estimation results in Question 5, test the null hypothesis that the mean test score in districts with low student-teacher ratio (STR<17)
is equal to the mean test score in districts with high student-teacher ratio (STR ≥ 17) . For this question, the standard error type in R has to be either
"stata" or "HC1". Choose the correct statement:
a. We do not reject the null hypothesis at the 10% significance level.
b. We do not reject the null hypothesis at the 5% significance level.
c. We do not reject the null hypothesis at the 1% significance level.
d. We do not reject the null hypothesis regardless of the significance level.
QUESTION 6 1 points   Save Answer
Using the estimation results in Question 5, test the null hypothesis that difference between the mean test score in districts with low student-teacher ratio
(STR<17) and the mean test score in districts with high student-teacher ratio (STR ≥ 17) is 7.9. For this question, the standard error type in R has
to be either "stata" or "HC1". Choose the correct statement:
a. We reject the null hypothesis at the 10% significance level.
b. We reject the null hypothesis at the 5% significance level.
c. We reject the null hypothesis at the 1% significance level.
d. We reject the null hypothesis regardless of the significance level.
QUESTION 7 1 points   Save Answer
Click Save and Submit to save and submit. Click Save All Answers to save all answers.