Introductory Econometrics
Assignment 1 (10%) Semester 2 2015
The due date for this assignment is 11 September.
You must read these guidelines/rules before proceeding with your assignment.
1. This assignment must be submitted to your tutor’s mailbox on Level5, Building H,
Caulfield Campus by 6:00pm on the above due date.
2. This assignment is worth 10% of the final mark for this unit.
3. Your assignment can be hand-written or typed and must be submitted in hard copy.
Submissions in electronic form or by email are not accepted.
4. Make sure the assignment COVER SHEET is on top of your assignment. You must complete
all sections of the cover sheet, including your tutorial day/time, and your tutor’s name. You
must also sign and date the cover sheet. If your assignment does not have a cover sheet it
will be returned to you unmarked. The cover sheet is available at URL:
http://www.buseco.monash.edu.au/student/forms/assessment-coversheet.doc
You may ask your tutor for any necessary information to complete the cover sheet.
5. This assignment must be no longer than 7 pages in total (including any graphs and tables.)
Any pages in excess of this length will be ignored in the marking process. Also, your work
must be clear and legible. If it cannot be read, or can be read only with difficulty, your
assignment will be returned to you unmarked.
6. ETF2100: The assignment will be marked out of 100 and the mark will be converted to a
mark out of 10 for the purpose of establishing your final mark in this unit.
7. ETF5910: The assignment will be marked out of 120 and the mark will be converted to a
mark out of 10 for the purpose of establishing your final mark in this unit.
8. Use A4 paper and make sure it is securely stapled at the top left corner.
9. You must keep a copy of your assignment. In addition, you must retain a copy of your
marked Assignment 1 until after the publication of the final results.
10. Late assignments, where approval for late submission has not been given, will be penalised
by deducting 25 out of 100 marks for each business day for up to four days, at which time a
mark of zero will be given.
11. If you are experiencing interference with your studies that is outside of your control, you
may be eligible to receive an assignment extension. Applications must be made by
completing the In-Semester Special Consideration Application Form found at the URL
http://www.buseco.monash.edu.au/student/exams/speccon.html
12. You should consult with your tutor before submitting this form to the lecturer. When it is
supported by your tutor, it should then be submitted to Professor Param Silvapule for
approval. Do not copy EViews output into your submission except for appropriate graphs.
Write estimated regression equations as equations, and tabulate other data you wish to
refer to in a reasonably compact form.
13. Your tutor will return assignments in class or during consultation time.
1
The data file includes crime and punishment data in the US for 1981 & 1987. The variable
CRIME is the crimes in the US, and PROARR is the probability of arrest, and it is measured
as the ratio of arrests to offences. The variable PRBCONV is the probability of conviction,
and it is measured as the ratio of conviction to arrests.
Question 1
Use the data crime_data_1987.xls and answer the questions from the US department of
criminology:
(i) Estimate the log-linear regression relating the log of the crime rate to the probability of an
arrest. That is estimate the simple linear regression model:
(1)
LCRIME = ß1 + ß 2 PROARR + e .
where LCRIME = log(crime rate)
Report the estimated regression results in the usual way.
(5 marks)
(ii) Comment on the goodness-of-fit of the model as measured by R2.
(5 marks)
(iii) Test the null hypothesis that there is no relationship between the crime rate and the
probability of arrest against the alternative that there is an inverse relationship between the
two variables. Use the level of significance 1%.
(15 marks)
(iv) If we increase the probability of arrest by 10%, what will be the effect on the crime rate?
(Hint: Compute the marginal effect for this log-linear model & interpret the marginal effect.) (5 marks)
(v) Compute the elasticity at the mean, if necessary. Interpret this quantity.
(5 marks)
(vi) Generate the residual plot. Does this plot indicate any pattern? If so, what can you infer
about the constant variance assumption about the error term?
(5 marks)
(vii) Test whether or not the error term is normally distributed.
(10 marks)
Question 2
An expert in criminology is saying that the crime rate cannot be reduced significantly by
increasing the probability of arrests, but it can be reduced by increasing conviction. Can you
test this claim?
Hint: You can answer the questions below:
(a) Now, use PRBCONV as the explanatory variable in model (1). Repeat the parts (i) to (vii).
(30 marks)
(b) Do you support the expert’s claim? Justify your answer by comparing the relevant results
of the two models.
(10 marks)
(c) Based on your findings, write a (maximum of) ½ page report to the department of
criminology on the relationship between the crime rate & the two variables and how the
department can reduce the crime rate.
(10 marks)
*Question 3 (This question is for ETF 5910 students only)
Use crime_data_1981.xls. Carry out the necessary analysis for the year 1981 & compare the
relationships between the crime rate and the explanatory variables across the two years. Write
a (maximum of) 3/4 page report. No need to conduct detail analysis again.
(20 marks)