- to acquire mastery of the tools of bivariate and multivariate quantitative data analysis.
- use single and multiple regression methods and some applications of generalized linear models (logistic regression and Poisson regression)
- understand and be able to use factorial analysis and classification techniques (also called cluster analysis)
- to be autonomous in the use of R, a free software for data analysis.
- Univariate analysis (reminders): to describe the data.
- Chi-square, relative risks, odds ratios: to analyze jointly two qualitative variables.
- T-Test, F-test and ANOVA: to test the relationships between a qualitative and a quantitative variable.
- Correlations, simple linear regression: to analyze jointly two quantitative variables
- Factorial analyses: principal component analysis (PCA) for quantitative variables and Multiple Correspondence Analysis (MCA) for qualitative variables: to construct indicators or identify 'latent' dimensions of all the variables analysed.
- Classification methods (Wald's hierarchical classification): to identify clusters of observation units or to develop typologies.
- Multiple linear regression and the generalized linear model (logistic regression and Poisson regression): to predict the value of a dependent variable, and identify its determinants.
Due to the COVID-19 crisis, the information in this section is particularly likely to change.The course is structured around lectures and practical work (see programme distributed in the first session and on Moodle). Participation in courses and partical sessions is essential. It is necessary to read chapters from the curriculum beforehand.
For the academic year 2020-2021, the course will be given in a comodal mode, with part of the students physically present, and part of the students at a distance.
Due to the COVID-19 crisis, the information in this section is particularly likely to change.
- A dispensatory test allowing students to evaluate their command of the R software is scheduled during the semester.
- Participation in three exercises associated with practical work is marked.
- The final evaluation is also based on a written exam during the exam session.
Inferface Rstudio: https://www.rstudio.com/
D.C. Howell, V. Yzerbyt, Y. Bestgen, and M. Rogier. Méthodes statistiques en sciences humaines. Série Internationale. De Boeck Supérieur, 2008.
- Les supports de cours, ainsi que la bibliographie complémentaire se trouvent sur la page Moodle du cours.
- G. Masuy-Stroobant and R. Costa, editors. Analyser les données en sciences sociales : De la préparation des données à l'analyse multivariée. P.I.E. Peter Lang, 2013. Disponible en bibliothèque et sur Moodle (certains chapitres)