The course provides a gentle introduction to the logic of quantitative statistical analysis in social and political sciences. By the end of the course, students who will attend classes will be able to:
· Develop a sound research design on social and political phenomena (i.e., a great emphasis will be placed on the formulation of hypotheses and on the use of data to test such hypotheses); · Perform descriptive and inferential multivariate analyses on R; · Master debates on the assumptions underlying the main statistical techniques social and political scientists use.
The programme is pondered and, if needed, adjusted depending on students' starting level. It aims at providing students with several analytical tools for making good empirical inferences in the realm of social sciences and training them in recognizing strengths and limitations of the evidences provided by others.
The first part of the course covers the basic logic of quantitative research design in social and political sciences and then refreshes univariate and bivariate analyses, Ordinary Least Squares (OLS) and its assumptions, as well as how to deal with violations of the basic linear model. Upon collective agreement, the second part will be dedicated to more advanced techniques for quantitative analysis (such as logit and probit regression models, time-series and panel-data analyses). Lectures will be based on hands-on material and will provide interactive learning experiences.
Expected learning outcomes
By the end of the course, students will be able to: -Perform descriptive and inferential multivariate analyses in R (i.e., multiple regression models, non-linear regression functions, limited dependent variables, time-series and panel data analyses). -Develop a sound research design on social and political phenomena (i.e., identify an original research question, formulate related hypotheses, find a suitable dataset, run the statistical analysis, interpret the results and discuss the limitations). -Discuss and defend the results of their research project through a catchy public presentation. -Master debates on the assumptions underlying the main statistical techniques social and political scientists use.
Lesson period: Third trimester
(In case of multiple editions, please check the period, as it may vary)
The first module by Prof. Negri (40 hrs) covers the basic logic of quantitative research design in social and political sciences and then refreshes univariate and bivariate analyses, Ordinary Least Squares (OLS) and its assumptions, as well as how to deal with violations of the basic linear model.
Upon collective agreement,the second module by Prof. Federica Genovese (20 hrs) will be dedicated to more advanced techniques for quantitative analysis (such as logit and probit regression models, time-series and panel-data analyses).
First module - Prof. Negri (40 hrs): 1. Quantitative Research Design in Social and Political Sciences 2. Refresh: Bivariate Regression Models 3. Refresh: Multiple Regression Models 4. Hypothesis Testing 5. Regression Diagnostics
Second module - Prof. Federica Genovese (20 hrs): 6. Non-linear Regression Functions - Quadratic and Interaction Models 7. Limited Dependent Variables 8. Time-Series and Panel Data 9. Drawing Graphs
Prerequisites for admission
Students have to be familiar with the topics treated in Data Analysis (Proff. Tommasi Chiara and Dotti Sani Giulia Maria) as well as to have a basic understanding of the software R as presented in Introduction to R (Prof. Vederico Vegetti). Instead, the mathematical requirements for the course are minimal: only a decent knowledge of algebra is assumed.
Lectures, hands on sessions in R, homework and team work.
The program is the same for attending and not-attending students. It covers: 1. John Fox and Sanford Weisberg, An R Companion to Applied Regression (3rd edition), SAGE Publications, 2019, Chapters 4, 5, 6, 8, 9. 2. Paul M. Kellstedt and Guy D. Whitten, The Fundamentals of Political Science Research (2nd edition), Cambridge University Press, (2009) 2013, Chapters 1, 2, 3, 4, 8, 9, 10, 11, 12; Chapters 5, 6, 7 (read only). 3. Thomas Brambor, William Roberts Clark, and Matt Golder (2006). Understanding Interaction Models: Improving Empirical Analyses. In Political Analysis 14(1): 63-82.
Assessement methods and criteria
Attendance is not compulsory, but it is warmly suggested.
Students who will attend classes will be assessed based on 1. their actively taking part to the class (i.e., attending all the sessions, reading the assigned chapters and solving class exercises); 2. individual homework (i.e., 3 individual home assignments); and 3. team work. As far as (3) is concerned, working in groups of three people (maximum), attending students will have to: 1. identify an original research question and formulate related hypotheses; 2. find a suitable dataset and develop a feasible research design to test the hypotheses; 3. run the statistical analysis; 4. present and discuss their results through a catchy presentation. The final grade will be computed as follows: (1) class participation will determine 15% of the grade; (2) the three home-assignments will determine 60%; (3) the team research project will determine 25% of the grade.
Non-attendant students will be evaluated through a written exam in the lab. They will be required to perform descriptive and inferential analyses on a dataset provided by the instructors by using R.