ECO374 Writing Assignments: Main Task
You will be given a part of a time series data.
Time periods 1, 2, , 180.
It is real monthly data.
You are asked to write a small essay in which you
1 Identify the pattern of the data.
2 Find the model that you believe capture the pattern best.
3 Estimate the model.
4 Make forecasting for 12 periods.
Imagine that you work at a consulting rm or at IMF…
Deadline: Noon, December 16, 2017
Submission through Blackboard, link will disappear after the deadline.
Data's identity will be revealed after everyone has submitted their essay.
Format Requirement
You are required to submit both the essay and the excel spread sheet containing your intermediate and nal results.
You shall report your main results in your essay too (your estimates for the model, your 12 period forecasts…)
The essay shall be 8{10 pages long
Including gures and tables.
Double space, 11pts.
Longer essay does not necessarily implies higher score.
Essay and spreadsheet must be submitted online through Portal.
Only ONE attempt is allowed.
The essay shall be in PDF format.
Format Requirement
Record your 12 periods forecasts in the last 12 periods of the data.
These 12 numbers are important for the evaluation.
Name your PDF and Excel le as \LASTNAME STUDENTID", e.g.
\WAN 1123455".
Grading
1. Basics: 12 points.
the essay shall be complete in content: contains your model, estimation and forecasting.
the essay shall be complete in structure: introduction, main part, and conclusion
Anyone who makes reasonable e ort is expected to receive all 12 points
2. Analytic Skills: 8 points.
3. Additional bonus: 5 points
Based on your point and interval forecasts.
Grading
2. Analytic Skills: 8 points
The essay shall be well structured and coherent.
Has strong and well supported arguments in model selection.
Whenever possible, carefully exams the assumptions required by the model.
Has clear explanation of the intermediate and nal statistical results.
Report both interval and point forecasts.
You will lose skill points if you use wrong formula, or your nal forecasts are inconsistent with the formula you use.
Other shinning points.
Grading
3. Additional bonus: 5 points
Only for those who submit the writing assignment on time.
For each student, a measure of forecasting error will be computed as
SSE = å (yˆT+t yT+t )2
t=1
yˆT+t is your prediction at time T + t.
yT+t is the real data at time T + t.
SSE will be computed up to 4 digitals behind the decimal, e.g.
10.2336.
Grading
3. Additional bonus: 5 points
Each students will receive a rank, according to their SSE value.
A student who has smallest SSE will have rank 1, second smallest will have rank 2 etc.
If there are N students submit assignment on time, and your rank is r , then your bonus points is
5 1 r N 1
Grading
Deadline: December 16, noon.
If you submit before deadline:
Your WA score = Basic + Skills + Performance
If you submit m days after the deadline:
Your WA score = (1 20% m) (Basic + Skills)
Some tips
Avoid making claims without arguments or statistical evidences.
You may want to include graphics to support or illustrate your points.
You may also want to divide your essay into small sections:
identi cation, estimation, forecasting…
Numbers and explanations are both necessary.
Methods: your choice. Could be any methods, as long as you provide a valid citation.
Whenever you cite any results in literature, state the reference clearly:
Paper title, author name(s), journal name, issue number, Vol.
number, year.
Only academic reference counts.
Academic Honesty
This is not a team projects: work on your own.
Suspected cases of plagiarism will be passed to university authority.
Outcome could be very serious.
