Accueil > Statistical reasoning

KOUT 2030 - STATISTICAL REASONING

Type d'enseignement : Seminar

Semester : Autumn 2017-2018

Number of hours : 24

Language of tuition : English

Pre-requisite

aucun

Course Description

This course is about the core notions of quantitative research for the social sciences, based on three fundamental blocks of knowledge: essential statistical concepts, survey data, and various forms of regression analysis. By design, this course will approach quantitative analysis through methods and examples taken from various branches of the social sciences, with some specific applications to international relations. We will focus on research design, as to make sure that we ask valid questions, based on sound hypotheses as well as reliable data, and draw correct inferences. Throughout the course, we will introduce and explain some essential statistical operations that can be used to that end. Finally, we will introduce statistical software and work through the procedures to produce statistical tests and visualizations of quantitative data. The emphasis of the course is set on conceptual understanding and statistical reasoning, and each session will apply statistical procedures to real data. Handbook chapters will be used to cover the statistical side of the course, while class sessions will focus on practical experience. No previous knowledge in any of these topics is required for taking the course, but some computer and Internet skills as well as a genuine interest in understanding why and how we use quantitative information to understand society will prove useful.

Teachers

  • BENOIT, Cyril (Attaché temporaire d'enseignement et de recherche)
  • BRIATTE, François (PhD Candidate - University of Grenoble)
  • DE LAEVER, Antonin (Doctorant Contractuel)
  • DESCHENES, Sarah (Doctorant)
  • JAROTSCHKIN, Alexandra (Etudiante doctorante)

Course validation

Students are expected to be regular and active participants in the course, and to complete required readings and exercises prior to class meetings. Course sessions start with a theoretical and practical introduction, after which students will train themselves to perform routine quantitative operations using Stata. The course sessions use a wide range of examples and exercises based on real data, and the course will regularly explore a selection of teaching dataset to familiarize students with survey data analysis. Students will be assessed on the basis of two draft papers and one final paper, for which they should provide replication material. The assignments and paper will all revolve around a single dataset and research question that students will gradually outline and apply throughout the semester. Expectations about coursework will be outlined at the first meeting and further detailed at several points of the course. Feel free to ask for additional guidance on what to read and how to structure your papers, yet do not wait for the last minute to do so, and read extensively from the course documentation. The grading policy for the course is 25 points for each report and 50 points for the final paper. Attendance to all sessions, which are all computer-based, is crucial to the course. Finally, students are asked to provide as much feedback on the course as they can.

Required reading

  • Agresti, A. and Finlay, M. 1997. Statistical Methods for the Social Sciences. 3rd ed. Prentice-Hall
  • Briatte, F. and Petev, I. 2012. Stata Guide. Online at http://f.briatte.org/teaching/quanti/.
  • Feinstein, C. H. and Thomas, M. 2002. Making History Count. Cambridge University Press

Additional required reading

  • Acock, A. 2008. A Gentle Introduction to Stata. 2nd ed. StataCorp
  • Booth, W. et al. 2003. The Craft of Research. 2nd ed. University of Chicago Press
  • Mitchell, M. 2004. A Visual Guide to Stata Graphics. StataCorp
  • Tufte, E. 2001. The Visual Display of Quantitative Information. Graphics Press