Quantitative data and methods: public administration and social sciences
Course description
Numbers are everywhere today: economy, health, politics, environment, education, sport, etc. Few spheres of our existence can claim to resist the contemporary quantitative deluge and the advent of Big Data. Admittedly, this situation is not totally new: figures have accompanied the development of modern science and the advent of contemporary forms of State. But the pace of change has accelerated dramatically in recent decades. And the profusion of data is not without raising thorny cognitive, ethical and political issues.
But what data exactly are we talking about? How is it produced? What are they used for? How can we use it in a reasoned way and understand it without making mistakes? With its theoretical and practical components, the course will present a broad overview of quantitative sources, reasoning and methods of analysis.
Drawing on elements of history and epistemology, students will be encouraged to take a reflective look at the tools of contemporary statistics and data science. As part of their dissertation, they will apply advanced techniques directly to contemporary data linked to a public issue or a social question.
- Topics covered include accounting modelling, the measurement of economic inequalities, quantitative methods in health, the statistical treatment of discrimination, the dangers associated with the use of digital data and, conversely, the opportunities for action they open up.
- The practical sessions include an introduction to the R software environment and present the methods of descriptive and factorial analysis, clustering and machine learning, and regression (linear and logistic).
The course provides an in-depth introduction to the quantitative skills required for jobs in economic administration and official statistics, whether at state or local authority level. Because of the social issues it covers, it will also be useful for students wishing to enter the field of non-governmental organisations and social science research.
Please note: access to individual data, which may be used in the dissertation, is subject to registration for the course being completed by mid-November 2024 at the latest.
Provisional programme
Session 1 (Thomas Amossé): what is quantification; dissertation themes; history and law of data; examples of use.
Session 2 (Camille Beaurepaire): R software; descriptive statistical methods; choice of dissertation topics.
Session 3 (Thomas Amossé): the economy and the environment taken into account; class struggle and economic inequalities.
Session 4 (Camille Beaurepaire): factorial analysis of data; application of descriptive methods to dissertation data.
Session 5 (Thomas Amossé, Camille Beaurepaire; Montrouge, INSEE library): the official statistical system; visit to the library and historical collections; documentary analysis for dissertations.
Session 6 (Camille Beaurepaire): clustering and machine learning; implementation on data for the dissertation.
Session 7 (Thomas Amossé): inequalities and discrimination (gender and race); statistical reasoning; tabular and graphical representations.
Session 8 (Camille Beaurepaire): linear and logistic regression methods; principles of data-visualisation; implementation on the data used for the dissertation.
Session 9 (Thomas Amossé): domination or emancipation, are data good or bad? From the Chinese social credit system to statactivism.
Session 10 (Thomas Amossé, Camille Beaurepaire): defence of dissertations.
- Teaching coordinator: Amossé Thomas
- Teaching coordinator: Beaurepaire Camille