The objective of the course is to present and discuss recent advances in applied quantitative methods with a point of departure in the challenges associated with the increased focus on causal inference in social sciences. The course aims at presenting different methods of dealing with these empirical challenges. We are targeting PhD students (and junior faculty, if space permits) with quantitative empirical projects in sociology, economics and political science. Students are expected to be competent in regression analysis, incl. the use of ordinary least squares, logistic, and similar types of regression techniques.
The series of single day Ph.D.-courses will in the spring semester cover the following theme:
Discussion of causality in social science – Using DAGs and regression discontinuity design
Lecturer: Anders Holm, Professor at University of Western Ontario
The aim of this one-day course is to understanding the logic of statistical control and omitted variables bias. This understanding is then framed in terms of potential outcomes framework. The identification of causal effects are then displayed in terms of causal graphs and causal graphs are linked to the potential framework. Finally, regression discontinuty (RD) is presented as one way of identifying causal effect. RD is also linked to causal graphs.
Date: 2nd of June, 2017, Kroghstræde 5, lokale 43/45
09.30-10.00: Registration, coffee and bread
15:15-16:30: Paper presentation (if anyone wishes to present)
Aalborg University, Kroghstræde 5, 9220 Aalborg Øst, lokale 43/45
participation = 1 ECTS, participation and paper presentation = 2 ECTS.
ECTS certificate is mailed about 2 weeks after the evaluation is closed.
Deadline for submission of paper (max. 5 pages) is Friday the 19th of May. The paper should focus on their PhD-project with special emphasis on the methodological challenges that they have and how that is related to the topic of the PhD-course.
Price and registration:
Participation is free for Ph.d.-students, other will have to pay for lunch (200 kr.)
Registration deadline: Friday the 26th of May to Anne Brauner Mikkelsen firstname.lastname@example.org
The organisers of the course are Professor Emerita Ruth Emerek, Associate Professor Sanne Lund Clement, Associate Professor Claus D. Hansen, Associate Professor Kristian Nielsen, and Associate Professor Rasmus Juul Møberg.
Regression and omitted variable bias:
Angrist, J. D. and J. Pischke (2015) chapter 2: “Regression” in Mastering Metrics – the path from cause to effect, Princeton university press.
Counterfactuals, potential outcomes and causal graphs
Morgan and Winship (2015), Chapter 2, “Counterfactuals and potential outcomes” in Counterfactuals and causal inference, Cambridge University Press.
Morgan and Winship, (2015), Chapter 3, “Causal graphs” in Counterfactuals and causal inference, Cambridge University Press.
Morgan and Winship, (2015), Chapter 4, “Models of causal exposure and identification criteria for conditioning estimators” in Counterfactuals and causal inference, Cambridge University Press.
Angrist and Pischke (2015) chapter 4, “Regression discontinuity Designs” in Mastering Metrics, Princeton University Press.
Suggestions for supplementary readings:
Freedman, D. A. (1991) “Statistical Models and Shoe Leather”, Sociological Methodology, Vol. 21, pp. 291-313.
Lieberson, S. (1985) Chapter 6: ”Control variables” in Making it count – the improvement of Social Research and Theory. University of California Press.