Applied Statistics for Political Scientists

Department of Politics and International Relations, University of Oxford

 

Developed by Steve Fisher, Stuart Soroka (McGill), and Sean Carey.

Instructors:    Sean Carey sean.carey@politics.ox.ac.uk

Steve Fisher steve.fisher@sociology.ox.ac.uk

                        Ken Macdonald  kenneth.macdonald@nuf.ox.ac.uk

Lauren McLaren lauren.mclaren@politics.ox.ac.uk

 

Class Assignments are here.

________________________________________

 

Readings and Assignments

Datasets and Codebooks

Links

The Nuffield College Library’s Taught Course Information website

________________________________________

 

Course Outline

 

(Man at Work)

 

 

This workshop is the practical, hands-on complement to the Statistical Methods for the Social Sciences lectures.  The aims are to (1) introduce topics and datasets from the empirical political science literature, and (2) familiarize students with STATA statistical software, data management, and methods of statistical analysis.

 

The course will survey a range of datasets and methods from across the discipline.  Examples are likely to include: regime type and the chances of war, government duration, political attitudes, and electoral behaviour.  Exercises and datasets will be studied in conjunction with related published literature.  This course has been adapted from one originally designed by Steve Fisher and Stuart Soroka.

 

Students taking the course for credit should remember that ALL this term's coursework is to be completed and bound together into a 'workbook' or 'portfolio' and handed in to the Politics Graduate Secretary, by the end of the working day Friday of 0th week Hilary Term 2003.

 

Week 1            Running and using STATA, including file management, data input, recoding and transforming data.  Graphical exploration and presentation of data including box-plots, histograms, scatter plots, pie charts and other methods. [Lecture Notes in pdf]

 

Week 2            Univariate statistics, including mean, mode, median, standard deviations, percentiles etc.  Tabulating and summarizing data.  Hypothesis testing for means (with known and unknown variance), difference in means and standard deviations of interval data.  Concepts of sampling and random data generation. [Lecture Notes in pdf]

 

Week 3            Categorical variables, contingency tables, notions and measures of association and independence, and tests for independence. [Lecture Notes in pdf]

 

Week 4            Linear regression. [Lecture Notes in pdf]

 

Week 5            Multiple linear regression, including tests for several partial regression coefficients, log transformations, interactions, standardized regression coefficients (betas). [Lecture Notes in pdf]

 

Week 6            Regression models for categorical variables, and weighting for complex survey data. [Lecture Notes in pdf]

 

Week 7            Interpretation of logit and probit models, linear regression diagnostics, and multicolinearity. [Lecture Notes in pdf] 

 

Week 8            Problems with and advantages of time series data.  Basic time series analysis common in political science.  Autoregressive distributed lag (ADL) models. [Lecture Notes in pdf]