|
|
University of Manitoba, Dept. of
Economics
|
|
Instructor: Norman E. Cameron,
Room 235 St. John's College.
Ph. 474-8114; this phone takes messages.
Email: cameron@ms.umanitoba.ca
Home page for 18.317: www.umanitoba.ca/faculties/arts/economics/cameron/317.html
Official office hours: Monday & Wednesday1.30-2.30. I am generally
available in my office other than just before this class and before and during
slots 2 and 27 (Thurs. 2.30-5.25).
Course
description: Probability, estimation and testing of economic models or hypotheses,
simple regression.
Pre-requisites: The only
prerequisites are 18.120 and high school algebra. Some students will have taken
introductory Statistics (such as 5.100 or 5.200, or more), and many students
will have taken or will be taking introductory calculus. Both will be useful,
but neither is necessary to do well in the course. The first half of the course
develops the concepts of probability, probability distributions, and inference
from statistical observations, training you to use and recognize each at work
around you. The second half introduces
the steps involved in Ordinary Least Squares (OLS) regression: specification of
an estimating equation, choice of data, use of econometric software, and
interpretation of results. The course
deals with OLS regression in much greater depth and detail than is done in
Statistics 5.200. 18.318 focuses almost exclusively on the use and misuse of
the OLS estimator.
Course
objectives: This course provides the background to excel in 18.318 or 61.308
(Intro to Econometrics, taught in the Agricultural Economics Dept.), both of
which deal thoroughly with multiple regression. If you have mastered the concepts of 18.317, then by December you
will be able:
- to identify and relate various types of probability and probability
distribution
- to update prior probabilities properly using Bayes’ Rule
- to convert probabilities to odds ratios and vice versa
- to identify the sampling variation or uncertainty of sample means and other
sample estimators
- to use sampling variation to test hypotheses and calculate confidence
intervals for estimates
- to evaluate different estimators using appropriate criteria
- to carry out all the steps of econometric regression using Eviews software
specifying a regression equation,
downloading data, generating estimated coefficients and test statistics,
producing relevant displays of
results, and interpreting what the results mean.
Course
methods: This is a hands-on course: you can only learn statistical and
econometric techniques by practicing them. These techniques are almost always
embedded in some software, which looks after the calculations. Most of what you have to do as an
econometrician is to learn (a) how to ask your software for what you want, and
then (b) how to explain and present the output the software gives you. We will
be practicing these steps frequently throughout the course. Some of the work on assignments will require
use of the computer software. Your final exam will involve some use of this
computer software. Almost all of your
empirical work in later courses will involve using this same software. So get an early start in working with the
software. Start with the Eviews
Basics section of the Help Menu, or the website for the course text, given
below. DO NOT delay this till just
before the final exam, even if you have a friend in the course who can use the
software for you up to that point!
Course
evaluation: Econometrics is a language with its own jargon, which you must become
comfortable using if you are to understand and explain what you will be doing
with numbers in economics. Such comfort
in using any language comes only with practice. To provide that practice, there will be a regular e-mail
component to the course, four assignments, a mid-term, and a final exam, as
well as many illustrations and discussions in class. The final grade for the course will be based on the results of all
but one (any one) of four individual assignments (30%), a mid-term test (20%),
a three-hour final exam (40%, using the course software, in the computer lab)
and sending in all but one (any one) of the bi-weekly emails within 72 hours of
its due date (10%). Since the mid-term test is included for feedback to you
rather than as summary evaluation by me, your mid-term mark will be raised to
your final exam mark if the latter is better.
The
emails are to be 150-word messages (not in attachments, please; they cause
problems) in which you try to explain and illustrate whichever course concept
is giving you the most difficulty. Your
explanation of that course concept need not be correct, but you must try to
explain and illustrate it. If you
understand the concept well to start with, then this component of the course is
easy. If you learn it in the process of
explaining it, the email has paid off before you even send it. If errors remain, you will get useful
feedback from me or the TA (your thinking coaches), that will help you get rid
of those errors in later uses. We will
not grade any of the emails, but we will note that each email has been
received, and provide feedback on any errors you may make. Take full advantage of this chance for
practice with feedback, without threat of evaluation. Due dates are every second Friday starting Sept. 17.
Text: We will use
Studenmund, Using Econometrics: A Practical Guide, 4th ed.,
(Addison Wesley) as the main text for 317.
It will also be suitable for 318. The computer software is EViews,
available on a CD in the bookstore at an excellent price, to be mounted on
whatever computer you use; the well-written User's Manual is built into the
program as a huge Help menu. The website for the Studenmund text
(http://occawlonline.pearsoned.com/bookbind/pubbooks/studenmund_awl/) contains
many useful data files, and a beginner’s guide to using the EViews software; I
have put this link on the website for this course. Other reference material will be noted during the course and
either put on reserve in the St. John’s College library or mounted on the
website.
Other:
You should acquaint yourself with the U of Manitoba policy on plagiarism and
cheating, laid out in the Undergraduate Calendar.
|
|
|