Advanced QuantitativePolitical Methodology POLS/CSSS 503 Continues the graduate sequence in quantitative political methodology. Focus on fitting, interpreting, and refining the linear regression model, gaining familiarity with statistical programming in R, developing clear and informative graphical representations of regression results, and understanding regression models in matrix form. Covers more advanced topics, including time series, panel data, and causal modeling, as time permits.

POLS/CSSS 503

Political Methodology

Offered every Spring at the
University of Washington
by various instructors

Syllabus

Spring 2014

Class meets:
Tuesdays 4:30-7:20 pm
Electrical Engineering 031

 TA: Carolina Johnson (UW Political Science)

Section meets:
F 1:30-3:20 pm
Savery 117

Lectures           PDFs of slides are best viewed in Adobe Acrobat, rather than in your browser.

Topic 1

Introduction to the Course and to R

Supplementary material: R code and data for the GDP example. R code and data from the fertility example.

You’ll find detailed instructions for downloading, installing, and learning my recommended software for quantitative social science here. Focus on steps 1.1 and 1.3 for now, and then, optionally, step 1.2.

Topic 2

Review of Matrix Algebra for Regression
and Regression and Graphics in R

Supplementary material: We will work through Kevin Quinn’s matrix algebra review (pdf). For a general review of the basic math we’ll need in the course, you can find the lecture notes for the CSSS Math Camp here.

R code and csv data for an example of how the base graphics package can create scatterplots and perform linear regression.

Topic 3

Linear Regression in Matrix Form and
Properties and Assumptions of Linear Regression

Supplementary material: You may find useful three review lectures on basic probability theory, discrete distributions, and continuous distributions.

Topic 4

Inference and Interpretation of Linear Regression

Supplementary material: Example code for estimating a linear regression, extracting confidence intervals for the parameters, and plotting fitted values with a confidence envelope.

Topic 5

Specification and Fitting in Linear Regression

Topic 6

Outliers and Robust Regression Techniques

Topic 7

Time Series: Stochastic Processes

Topic 8

Models of Stationary and Non-Stationary Time Series

Supplementary material: Example code and csv data for estimating and interpreting ARMA models in R, and example code for estimating and interpreting ARIMA and ECM models. You will also need this helper function for plotting counterfactual time series using R base graphics.

Topic 9

Introduction to Panel Data Analysis

Supplementary material: Example code and csv data for estimating and interpreting panel ARIMA models in R. For the curious, the R script used to construct the example plots in the first half of this lecture is here.

Student Assignments

Problem Set 1

Due Tuesday, 15 April, in class

Supplementary material: Data for problem 1 in comma-separated variable format.

Problem Set 2

Due Friday, 25 April, in section

Supplementary material: Data for problem 1 in comma-separated variable format.

Problem Set 3

Due Tuesday, 6 May, in class

Supplementary material: Five R script templates for simulation of the performance of linear regression with different kinds of data: when the Gauss-Markov assumptions apply; when there is an omitted variable; when there is selection on the response variable; when there is heteroskedasticity; and when there is autocorrelation in the response variable.

Problem Set 4

Due Tuesday, 20 May, in class

Supplementary material: Data for problem 2 in comma-separated variable format.

Problem Set 5 (Optional)

Due Friday, 6 June, in section

Supplementary material: Data for problems 1. Data for problem 2. Data for problem 3. (All data in comma-separated variable format.)

Final Paper

Due Monday, 9 June, at 3:00 PM, in my Gowen mailbox

See the syllabus for paper requirements, and see my guidelines and recommendations for quantitative research papers.

Labs

Lab 1

Introduction to R

Lab 1 code and a page of links to R and LaTeX Resources.

Lab 2

Beginning with data: reading in, exploring, and even regressing

Lab 2 code and data for Example 1 and Example 2.

Lab 3

More practice with regression and plotting

Lab 3 code. Data is the same rossoil data as Example 1 in Lab 2.

Lab 4

Useful things, largely loops and simulation

Lab 4 code. Data for the first set of examples. Data for the second (model simulation) example.

Lab 5

More on models and interpretation, simulation in detail

Lab 5 code. Data for the first set of examples. It's the same rossoil data we've been using for weeks.

Lab 6

Miscellaneous, transformation, and heteroskedasticity

Lab 6 code. This will use two datasets: the same rossoil data we've been using and the democracy data from earlier.

Lab 6+

Intro to LaTeX workshop

Lab 7

Intro to logit and functions (and a bit of outliers)

Lab 7 code. This will use two datasets: the same rossoil data we've been using and NES data for logit examples.

Lab 8

Intro to simcf/tile (including for logit)

Here is Lab 8 code. This lab will use the same datasets as last week's lab: the rossoil and NES data. We will also access the data for HW4 online in the script.

Lab 8

Very brief intro to useful functions like apply() and plyr package

Here is Lab 9 code. All data will be called within script.