University of Washington, Spring 2015
Teaching Team
Professor  Jeffrey Arnold  jrnold@uw.edu 
TAs  Sergio GarcíaRios  sigarcia@uw.edu 
Carolina Johnson  csjohns@uw.edu 
Class Meetings
Class  Tues  4:307:20 pm  SAV 158 
Lab  Fri  1:303:30 pm  SAV 117 
Office Hours
Jeffrey Arnold  Th 3:00–5:00 pm  Smith 221B 
Carolina Johnson  Tu 3:30–4:30 pm  SAV 119 
Sergio GarcíaRios  F 12:30–1:30 pm  SAV 119 
Overview and Class Goals
This course continues the graduate sequence in quantitative political methodology, focused particularly on fitting, interpreting, and refining the linear regression model.
Our agenda includes gaining familiarity with statistical programming via the popular R environment, developing clear and informative graphical representations of regression results, and understanding regression models in matrix form.
Prerequisites
It is desirable for students to have taken the introductory course in the sequence (Political Science 501), but any prior course on basic social statistics and linear regression should suffice.
Assessment and Evaluation
Problem Sets
Problem sets assigned weekly or biweekly. These problem sets will include programming problems with an emphasis on writing understandable, reproducible code.
The assignments and due dates will be distributed during the quarter.
Assignments will be both submitted digitally through Canvas at the due date and a paper copy to the TA at the next lab section.
Final Paper
A 15–20 page original report on an original quantitative analysis or replicationandextension of a published article. The quantitative analysis should be conducted in R and reproducible. Students may work in pairs on the final paper with instructor permission.
The final paper is due on June 9, 2015 15:00 PDT.
For details, see the Guidelines for the Final Papers.
Email & Canvas:
The teaching team will send announcements regularly by email.
Any nonpersonal questions related to the material in the course should be posted as a Canvas discussion. Reserve email for personal or administrative matters. Before posting, check that the question has not been asked and answered already.
It is often more efficient to answer questions in person, so try to ask them attend office hours.
Resources
There are a couple places on campus that you can go to get additional statistical conulting
 CSSCR has a dropin statistical consulting center in Savery 119. They provide consulting on statistical software, e.g. R. http://csscr.washington.edu/consulting.html
 CSSS Statistical Consulting provides general statistical consulting (questions about your research project). You can find their hours and locations on thier site.
Texts
Required
 Fox, John. 2008. Applied Regression Analysis and Generalized Linear Models. 2nd edition. Los Angeles: SAGE Publications, Inc.
 Gelman, Andrew, and Jennifer Hill. 2006. Data Analysis Using Regression and Multilevel/Hierarchical Models. 1st edition. Cambridge ; New York: Cambridge University Press.
 Angrist, Joshua D., and JörnSteffen Pischke. 2009. Mostly Harmless Econometrics: An Empiricist’s Companion. 1st edition. Princeton: Princeton University Press. (UW library eBook available)
Optional
 Zuur, Alain, Elena N. Ieno, and Erik Meesters. 2009. A Beginner’s Guide to R. Springer. (UW library eBook available)
Schedule and Topics
Warning: The uncertainty on this schedule is high. Given that this is my first time teaching this course, I expect many adjustments to be made.
Week 1: Introduction to 503 and R, Math Review
Tuesday, March 31
 Deck 1
 Inclass R Markdown example analysis: Rmd source, html output
 Readings
 Gelman and Hill, Ch 2
 Fox, Ch 2, 3 (review)
Friday, April 3
 Lab document: .Rmd, html
 Data: gapminder.csv

Readings:
 Data Camp “Introduction to R”, https://www.datacamp.com/courses/freeintroductiontor
 Wickham, Hadley. 2010. ``A Layered Grammar of Graphics.’’ Journal of Computational and Graphical Statistics 19(1): 3–28. http://dx.doi.org/10.1198/jcgs.2009.07098
Week 2: Assumptions & Properties of the Linear Regression Model, Part I
Tuesday, April 7
 Deck 2
 Readings:
 Matrix algebra readings. Read any of the following
 Moore, Will H., and David A. Siegel. 2013. A Mathematics Course for Political and Social Research. 1st edition. Princeton, NJ: Princeton University Press, Chapter 12 (on Canvas).
 Kevin Quinn’s matrix algebra handout
 CSSS Math Camp Lectures Section 4
 Fox, Ch. 5, 9.1–9.2
 Matrix algebra readings. Read any of the following
Friday, April 10

Readings:
 Hadley Wickham, Introduction to dplyr
Week 3: Assumptions & Properties of the Linear Regression Model, Part II
Tuesday, April 14
 Deck 3
 Sampling Distribution of linear regression example
 Multiple regression coefficient anatomy

Readings:
 Fox, Ch. 6, 9.3
Friday, April 17
Problem Set 1 Due
 Lab document: .Rmd, html

Readings
 Wickham, Hadley. 2014. ``Tidy Data.’’ Journal of Statistical Software 59(10). http://www.jstatsoft.org/v59/i10/.
 Hadley Wickham, tidyr vignette
Week 4: Statistical Inference / Interpretation of the Linear Model
Tuesday, April 21
Friday, April 24
 Lab document: .Rmd, html
 Data: ross_2012.csv

Readings
Week 5: Model Fitting and Data Transformation
Tuesday, Apr 28
Problem Set 2 Due

More on pvalues and significance testing
 Nuzzo, Regina. 2014. ``Scientific Method: Statistical Errors.’’ Nature 506(7487): 150–52. http://www.nature.com/doifinder/10.1038/506150a

Omitted Variable Bias
 Omitted Variable Bias handout
 Fox, Ch. 6.3
 Agrist and Pischke, Ch. 3.2.2

Heteroskedasticity and misspecification
 OLS Residuals deck
 Fox, Ch 12.1–12.3
 Angrist and Pischke, Ch. 8
 King, Gary, and Margaret E. Roberts. 2015. ``How Robust Standard Errors Expose Methodological Problems They Do Not Fix, and What to Do About It.’’ Political Analysis 23(2): 159–79. http://pan.oxfordjournals.org/content/23/2/159

Mispecification, Transforming Covariates
 See Christopher Adolph’s slides from 503 last year http://faculty.washington.edu/cadolph/503/topic5.pw.pdf
 Fox, Ch. 4, Ch 17.1–17.3
 Gelman and Hill, Ch. 4.1–4.3, 4.54.6
 Brambor, Thomas, William Roberts Clark, and Matt Golder. 2006. “Understanding Interaction Models: Improving Empirical Analyses.” Political Analysis 14(1): 63–82. http://pan.oxfordjournals.org/content/14/1/63
Friday, May 1
Week 6: Measurement Error, Transformations
Tuesday, May 5
We covered substantive vs. statistical influence; transformations; and measurement error. You will need to cover collinearity, unusual and influential data, and robust regression on your own.
 The perils of stargazing:
 Christopher Adolph, “Inference and Interpretation of Linear Regression”, POLS 503, Spring 2014. http://faculty.washington.edu/cadolph/503/topic4.pw.pdf#page=77, pg. 77–.

Transformations of Variables
 Lecture deck: .pdf

Measurement error
 Lecture deck: .pdf
 Fox. Ch. 6.4

Collinearity
 Fox Ch 13 (13.1, skim the rest)
Friday, May 8
Week 7: Interpretation, Model Selection
Tuesday, May 12
Problem Set 3 Due
 Deck on Model Specification and Fit
 Life Expectancy Example: .html, .Rmd

Interpretation of regression
 King, Gary, Michael Tomz, and Jason Wittenberg. 2000. ``Making the Most of Statistical Analyses: Improving Interpretation and Presentation." American Journal of Political Science 44(2): 347–61. http://www.jstor.org/stable/2669316.

Bootstrapping
 Fox, Ch 21

Model Selection and CrossValidation
 Fox, Ch 22
Friday, May 15
Week 8: Causal Inference I
Tuesday, May 19

Unusual and Influential Data and Robust Regression
 Worked Example Rmd, html
 Christopher Adolph, “Outliers and Robust Regression Techniques”, POLS 503, Spring 2014. http://faculty.washington.edu/cadolph/503/topic6.pw.pdf

Readings
 Fox, Ch 11, 19

Missing Data

Miscellaneous Thoughts on the state of quantitative analysis in political science. We did not talk about this, but these are good readings
 Schrodt, Philip A. 2014. ``Seven Deadly Sins of Contemporary Quantitative Political Analysis.’’ Journal of Peace Research 51(2): 287–300. http://jpr.sagepub.com/content/51/2/287
 Achen, Christopher H. 2002. ``Toward a New Political Methodology: Microfoundations and ART.’’ Annual Review of Political Science 5(1): 423–50. http://dx.doi.org/10.1146/annurev.polisci.5.112801.080943. Read the part on a “Rule of Three”; skim other parts.
 Achen, Christopher H. 2005. ``Let’s Put GarbageCan Regressions and GarbageCan Probits Where They Belong.’’ Conflict Management and Peace Science 22(4): 327–39. http://cmp.sagepub.com/content/22/4/327 skim
Friday, May 22
Week 9: Causal Inference II
Tuesday, May 26
Problem Set 4 Due

Limited Dependent Variables
 Slides on the Linear Probability Model and Logit pdf

Potential outcomes framework, regression, matching
 Slides on Casual Inference pdf

Readings
 Gelman and Hill, Ch 9–10.
 Angrist and Pischke, Ch 1–3.
 Angrist, Joshua D., and JornSteffen Pischke. 2010. ``The Credibility Revolution in Empirical Economics: How Better Research Design Is Taking the Con out of Econometrics.’’ Journal of Economic Perspectives 24(2): 3–30. http://www.aeaweb.org/articles.php?doi=10.1257/jep.24.2.3 For background reading
Friday, May 29
Week 10: Panel Data
Tuesday, June 2
 Slides on Panel data: pdf
 Example with R code: html, Rmd

Readings
 Angrist and Pischke, Ch. 5, 8.2
 Peter Kennedy, A Guide to Econometrics, 6th edition, Chapter 18, “Panel Data”. On Canvas.
 Wooldridge, Econometric Analysis of Cross Section and Panel Data, Chapter 10, “Basic Linear Unobserved Effects Panel Data Models”. On Canvas.
 Beck, N., & Katz, J. N. (2011). Modeling Dynamics in TimeSeries–CrossSection Political Economy Data. Annual Review of Political Science, 14(1), 331–352. http://doi.org/10.1146/annurevpolisci071510103222
Friday, June 5
Open office hours to answer questions about your papers
Syllabus derived from Christopher Adolph. (Spring 2014). POLS/CSSS 503: Advanced Quantitative Political Methodology [Syllabus]. University of Washington. http://faculty.washington.edu/cadolph/503/503.pdf CCBYSA.