Applied Regression Analysis 2
Applied Regression Analysis II (ARA2) builds on ARA1. ARA1 focused on cross-sectional analysis in R and sought to explore practical problems such as interpretation, visualization, missing data, interactions and logistic regression. By contrast, ARA2 shows why applied regression analysis requires simulation, adds temporal dynamics and explores panel data models, touching upon topics as instrumental variables, multilevel modeling, missing data in panel data and other issues.
Because students gained a hands-on approach in ARA1, they will now be required to present their own research at the end of the semester using the techniques discussed in ARA1 & 2, or other techniques they wish to present to the class.
This course covers the basics of simulation to check model fit and for model interpretation; the basics of text analysis; basic models for time series analysis and time series cross sections (aka panel data). We will emphasize model interpretation, simulation and presentation as well as dynamic modeling and forecasting. Panel data and time series are the bread and butter of public policy analysis and a standard technique in other fields like political science, political economy, sociology and economics. It is therefore crucial to understand how these models work.
More specifically, students will gain the following skills:
- Be able to interpret and explore dynamic models in R;
- Assess research of their peers and articles in leading journals;
- Discuss the policy relevance of econometric analysis;
- Get exposed to advanced statistical techniques in regression analysis;
- Become fluent in the statistical programming language R;
- Be able to conduct independent quantitative research;
- Get acquainted with software surrounding statistical analysis (e.g. LaTeX).
Students will be asked to perform three tasks on top of preparing for classes by reading and analyzing code. First, they will have to write a small critical essay on a chosen article (replicate the findings and assess whether the authors did a good job, 1000 words maximum) [30%]. Second, they will have to discuss the project of one of their colleagues by pointing out flaws and, most importantly, potential solutions in a written report (maximum 500 words) [30%]. Third and finally, they will have to submit their term project on April 7 (maximum 2000 words) [40%].