Introduction to Experimental Research in the Social Sciences

Social scientists conducting empirical research are more and more often confronted with questions regarding the validity of their causal claims. Whereas in many cases presenting some observational data is enough to validate the plausibility of a theoretical idea, testing causal hypotheses requires different approaches. Experiments are one of the most powerful tools to reach this goal, and they are becoming increasingly popular in social scientific research as well as in other fields where decisions are evidence-based.

This course aims at providing a basic theoretical and methodological foundation to political science students interested in experimental research. By the end of the week students will have a sound understanding of the logic of experimental design, and will be aware of some important issues that may arise in the process. Moreover, the practical sessions and readings will encourage them to use their inventiveness to develop original and potentially successful experimental studies.

The first part of the course will cover basic theoretical and methodological principles related to experimental research in social and political science. Day 1 will work as a general introduction to the course. We will review the history of experiments in social and political science, with a special focus on causality and inference. Moreover, we will discuss on how to convert a researchable idea into an experimental study, from both a theoretical and a practical perspective. In a practical session, we will brainstorm around a given research question, discussing ways to investigate it using experiments.

Day 2 will cover some essential elements on the logic of experimental design, like control and randomization, reduction of error, randomized block, and factorial designs. Day 3 will focus more on methodological and practical concerns: we will discuss internal and external validity, measurement, and number of observations. On both days 2 and 3, we will leave 1 hour at the end of the sessions to develop the research ideas introduced on day 1 in light of the concepts discussed during the lecture.

Day 4 will be an overview of some of the experimental designs commonly used in social and political science, like survey experiments, field experiments, natural experiments and quasi experiments.

Day 5 will bring some stats into the picture, offering an overview of the techniques used to analyze experimental data. The session will be split in 2, with half session consisting of a classic frontal lecture on statistical methods, and half session consisting of practical data analysis using the software R.

At the end of the course, students will know have of the theoretical and empirical tools to design a good quality experiment and to analyze the data produced by it. Moreover, they will have collected a number of examples and ideas of possible experimental designs, which will serve as inspiration for their own research work.

Basic concepts of empirical research and causal inference