First 4-Week Session
■Introduction to Applied Bayesian Modeling for the Social Sciences
This course introduces the basic theoretical and applied principles of Bayesian statistical analysis in a manner geared toward students in the social sciences. The Bayesian paradigm is particularly useful for the type of data that social scientists encounter given its recognition of the mobility of population parameters, its ability to incorporate information from prior research, and its ability to update estimates as new data are observed. The course will begin with a discussion of the strengths of the Bayesian approach for social science data and the philosophical differences between Bayesian and frequentist analyses. Next, the course will cover the theoretical underpinnings of Bayesian modeling and provide a brief introduction to the primary estimation algorithms. The bulk of the course will focus on estimating and interpreting Bayesian models from an applied perspective. Students will be introduced to the Bayesian forms of the standard statistical models taught in regression and MLE courses (i.e., normal, logit/probit, Poisson, etc.).
■Regression Analysis III: Advanced Methods
Linear regression is the workhorse of social science methodology. Its relative robustness and easy interpretation are but two of the reasons that it is generally the first and frequently the last stop on the way to characterizing empirical relationships among observed variables. This course will extend the basic linear model framework in a number of directions in an attempt to fix potential problems in the analysis before they arise. The course takes a modern, data-analytic approach to regression emphasizing graphical tools to aid interpretation and presentation of results. The course moves through both low- and high-tech diagnostics for and solutions to common obstacles for estimation of linear models, namely non-linearity, outlying observations, and dependent data. We will also spend some time talking about causal inference, missing data, and model selection (including multi-model inference).
Second 4-Week Session
■Simultaneous Equation Models
This course centers on simultaneous equation models -- models of more than one equation, to account for more than one dependent variable -- formerly called "causal models." The workshop will focus on linear models of the standard econometric type, continuing on to nonlinear models, models with discrete dependent variables, and models with measurement error ("covariance structure models") as time permits. The course will cover the nature of simultaneous equation models, their parameters, and the "effects" they imply; the assumptions under which one would customarily analyze them; the identification problem and criteria for identifiability; and simultaneous equations estimators such as two- and three-stage least squares, and limited- and full-information maximum likelihood.
■Time Series Analysis
This four-week workshop begins by focusing on the autoregressive and moving average components of time series, and then turns to estimation of univariate time series models using the Box-Jenkins approach. Intervention analysis and more general transfer function models build on this tradition, often referred to as the statistical analysis of time series. The course then focuses on the econometric (regression) analysis of time series, historically quite distinct from the statistical tradition. In recent years, regression analysis has borrowed much from the statistical tradition, and the connections between the two are important for understanding how social scientists should analyze time series data. Analysis of integrated time series, including unit root econometrics and error correction models, focuses on recent econometric advances in dealing with nonstationary data.