2007-2008 ITV Course Schedule
Fall:
"Nonparametric
and Robust Estimation" - Luke Keele
Fridays: 11:00- 1:00 CST
Email:
keele.4@polisci.osu.edu
"Potential Outcomes
Inference" - Jake Bowers
Wednesdays: 1:30- 3:30 CST
Email:
jwbowers@umich.edu
Spring:
"Time
Series" - Jan Box-Steffensmeier, John Freeman and Jon Pevehouse
Fridays: 11:00-1:00 CST
Email: steffensmeier.2@osu.edu
Email:
pevehouse@polisci.wisc.edu
Course Descriptions
Nonparametric Robust Estimation
Instructor: Luke Keele, Ohio State University
Email:
keele.4@polisci.osu.edu
Times: 11:00am - 1:00pm CST, Fridays
Dates: Sept 21, Sept 28, Oct 5th, Oct 19th, Oct 26th, Nov 2nd, and Nov 9th (all Fridays)
Description:
This course is designed to introduce graduate students to a
variety of advanced and computationally intensive methods that are
starting to be used regularly in the field. The class will
familiarize students with such topics as bootstrapping,
non-parametric and semi-parametric estimation, and robust
estimation. The topics in this class are not standard fare in the
typical methods sequence, but are appearing with increasing
frequency in applied work, and in many cases should be used more
often. The course will be run more along the lines of a workshop,
and it is hoped that there will be extensive interaction during
class as we review the methods covered here.
Potential Outcomes Inference
Instructor: Jake Bowers, University of Illinois
Email:
jwbowers@umich.edu
Times: 1:30pm - 3:30pm CST, Wednesdays
Dates: Sept 26, Oct 3, Oct 10, Break, Oct 24, Oct 31,Nov 7, Nov 14 (all Wednesdays)
Description:
The potential outcomes approach to causal inference (invented
by Neyman, developed by Rubin) emphasizes research design and
conceptual definition of a causal estimand over concerns with
properties of estimators (e.g. consistency, unbiasedness) or
over concerns about the chance processes that may have led to a
particular observed outcome (e.g. frequentist hypothesis testing
or bayesian posterior inference). Starting from an understanding
of how to think abut causal effects one is soon led to (1)
choose particular data analytic techniques (such as propensity
scores and matching) over others (such as linear models with
long lists of "control" variables) and (2) understand old
techniques in new light (such as understanding what assumptions
are required for us to believe $\hat{\beta_{OLS}}$ tells us
something about causality).
In this course, we'll start by learning what a potential outcome
is and how it can structure our thinking about causal
inference. Then we'll move on to learn about tools for making
such inferences, such as matching, propensity scores, and OLS
regression. As we learn about these different techniques for
"controlling for confounds" and for estimating causal effects,
we'll also learn about how to answer the question "Could this
effect be due to chance?" using "standard" hypothesis testing
(i.e. Neyman-style model based inference), as well as Bayesian
predictive inference, and Fisher-style permutation or
randomization inference.
Assignments can be completed in any programming language that is
scriptable, as long as the code used to complete the assignments
is submitted in a form that the teaching staff can submit for
interpretation to a program all at once (i.e. a ".do" file, or
an ".R" or ".Rnw", or even a ".c" file). The instructor will
exclusively use R and will provide R code for all of the
examples discussed in class.
Time Series
Instructors: Jan Box-Steffensmeier, Ohio State University
John Freeman, University of Minnesota
Jon Pevehouse, University of Wisconsin-Madison
Emails: steffensmeier.2@osu.edu
pevehouse@polisci.wisc.edu
Times: 11:00am - 1:00pm CST, Fridays
Dates: Jan 25, Feb 1, Feb 8, Feb 15, Feb 22, Feb 29, Mar 7, Mar 14, Apr 4, Apr 11, Apr 18, Apr 25, May 2 (all Fridays)
Description:
This course considers statistical techniques to evaluate social
processes occurring through time. The course introduces students
to time series methods and to the applications of these methods
in political science. After a brief review of the calculus of
finite differences and other estimation techniques, we study
stationary ARMA models. In the next section of the course, we
examine a number of important topics in time series analysis
including "reduced form" methods (granger causality and vector
autogression), unit root tests, near-integration, fractional
integration, cointegration, and error correction models. Time
series regression is also discussed (including pooling
cross-sectional and time series data). We learn not only how to
construct these models but also how to use them in policy
analysis.
We expect students to have a firm grounding in probability and
regression analysis and to bring to the course some interesting
questions about the dynamics of political processes. The
emphasis throughout the course will be on application, rather
than on statistical theory. However, the focus of most lectures
will be statistical theory. Homework will revolve as much as
possible around the time series you are interested in
understanding. To that end, students will need to gather time
serial data for analysis during the first week of class (this
data need not be used throughout the term, though that would
make your life easier). The length of the series should be at
least 40 time points; longer series are better than shorter
ones.
This is the first part of a fourteen-week seminar team-taught by
Professors John Freeman, Janet Box-Steffensmeier, and Jon
Pevehouse. Students are strongly encouraged to take both parts
of the course.
Back to Courses Page.
