NOTES:

Problem Sets, Exams, and Grading Policy
Grading: As per departmental guidelines, expect the following distribution of letter grades: A (25%), B
(45%), C (17.5%), D/F (12.5%). There are four graded components:

• Problem Sets: 15% (only 3 highest count)
• Statistical Assignment: 10%
• R Participation: 5%
• Midterm: 30%
• Final Exam: 40%

Problem Sets
There will be 4 problem sets assigned during this course. Every student has to hand in
his/her own work and group submissions are not allowed. Only the 3 highest marks on problem sets will be
counted toward grade, each one counting for 5% each (so lowest grade on these will be ‘thrown away’).

Statistical Assignment
This class will act as a very soft introduction to working with data and statistical
software. I will therefore assign one statistical assignment that will ask you to work with real data to estimate
either the impact of pollution on some economic outcome or the impact of some event (either manmade or
natural) on pollution (your choice). A data set will be provided; your job will be to link this data set to one
you create and report basic summary statistics and estimate a simple one variable regression to estimate
your relationship of interest. The result from the regression need not go the “expected” direction; you just
simply need to run it and explain what the coefficient means. The lectures will provide all the code you need
for the project (see below); you simply have to alter the code to fit your setting. You are allowed to work in
groups of 2-3 students if you choose.

R Participation
I understand that many of you will have very limited to no knowledge of R. I will therefore
devote 15-30 minutes of our lecture in weeks 3-6 to a coding “camp” to learn R. Please bring your laptops
to these classes with R installed and the data (which will be provided to you in Canvas) loaded so that you
can follow along with the code. A simple exercise will be given for you to complete in the last 5 minutes of
the class (e.g., fill in the final two lines of code). You will be asked to email the code to complete this simple
exercise to the RA; your mark is simply based on submitting a (somewhat) reasonable looking few lines of
code. These marks are just participation marks; you must complete 3 of them (out of 4) to get the full 5%
R participation grade. (This is so you can miss one of these four classes without consequence.)

Exams
We will have one midterm for this course which will be held in class on October 23. The midterm
will take 90 minutes. The final exam date is set by the university and will be held on TBD. The exams are
closed book exams and no notes will be allowed. A non-graphing calculator is allowed. Absolutely no other
materials may be used. Exam material will cover topics covered in lectures up to the midterm date as well
as concepts and ideas introduced in problem sets. Exams are assumed to be cumulative, but obviously there
will be an emphasis on recently covered material. You will be given guidelines on the topic coverage in the
‘Review Class’ before each exam.
Practice Exams: I will provide two practice exams in advance of each exam. Expect these to be similar
to the actual exam.

Regrading Policy
We will try very hard to grade exams as accurately as possible. However, errors do
happen. There are two types of errors: (i) your mark was incorrecly tabulated (i.e., the marker made an
addition error) and (ii) you felt that a question was marked unfairly. For the first type of error, please bring
your exam to myself within a week of it being returned and ask for the tabulation error to be fixed. For
the second type of error, you must make a request in writing for a regrade, explaining the issue. While
inquiries about how a question was graded or marked are fine, any questions about regrading (besides an
aforementioned tabulation error) that are not submitted in writing will be ignored. Some details:

• Regrade requests must be made in writing and submitted to myself
• Request must be made within one week from when exams are returned
• The entire exam will be regraded, meaning scores may go up as well as down
• There are absolutely no other options for having your mark on an exam question changed

Late Policy
As tutorials will go over answers in problem sets and you are given one “mulligan” on them,
late problem sets will not be accepted. Late statistical assignments will be punished by a 20% per day late
penalty. Do not hand in assignments late!