KSOD 2220 - Causal analysis in sociology

This course is a non-technical introduction to statistical methods at the heart of the “credibility revolution” in the Social Sciences: experiments, quasi-experiments, regression discontinuity design, difference-in-differences, instrumental variables and panel regressions. These methods enable to make more convincing causal claims.

The course will be as follows. For each notion, a lecture will develop the key concepts and provide basic intuitions on the functioning of the method. It will be followed the week after by an implementation of the method with real data where we replicate some article’s tables.

To follow this course, students must have a basic command of R software and of statistical notions such as statistical tests (Student and Chi-square tests), OLS and logistic regressions.

• Econometric Reminder. Lecture

Reading: Herpin, Nicolas. 2003. “La taille des hommes : son incidence sur la vie en couple et la carrière professionnelle”, Economie et Statistique, 361, p. 71-90.

• Econometric Reminder. Implementation

Replication: Herpin, Nicolas. 2003. “La taille des hommes : son incidence sur la vie en couple et la carrière professionnelle”, Economie et Statistique, 361, p. 71-90.

• Experimental methods (natural experiments, difference-in-differences, regression discontinuity design). Lecture

Reading: Mize, Trenton D., and Bianca Manago. 2018. “Precarious sexuality: How men and women are differentially categorized for similar sexual behavior.” American Sociological Review 83(2): 305-330.

• Experimental methods (natural experiments, difference-in-differences, regression discontinuity design). Implementation

Reading and replication: Godechot, Olivier, 2016. The Chance of Influence: A Natural Experiment on the Role of Social Capital in Faculty Recruitment. Social Networks, vol. 46, p. 61-75.

• Instrumental variables. Lecture

Reading: Goux, Dominique and Eric Maurin. Close neighbours matter: Neighbourhood effects on early performance at school. The Economic Journal, 2007, vol. 117, no 523, p. 1193-1215.

• Instrumental variables. Implementation

Replication: Goux, Dominique and Eric Maurin. Close neighbours matter: Neighbourhood effects on early performance at school. The Economic Journal, 2007, vol. 117, no 523, p. 1193-1215.

• Panel regression. Lecture

Reading: WALDFOGEL, Jane. The effect of children on women's wages. American sociological review, 1997, Vol. 62, No. 2 (Apr., 1997), p. 209-217.

• November 22nd 2022.12h30. Room H003. Panel regression. Implementation

Replication: WALDFOGEL, Jane. The effect of children on women's wages. American sociological review, 1997, Vol. 62, No. 2 (Apr., 1997), p. 209-217.

• Mixing it all (Complements to be defined with students: Diff-in-Diff and Lag Dependent Variable)

In order to validate this class, students must write a 2000 words (maximum) statistical memo (references and tables not included) where they apply one of the causal methods thought in class (weeks 3 to 9).