Dates: 17 - 21 August 2015
Standard rate: £1500
Academic rate: £890
Professor Brigitte Le Roux (Université Paris Descartes)
Professor Johs Hjellbrekke (University of Bergen)
Professor Mike Savage (Department of Sociology, LSE)
Dr Daniel Laurison (Department of Sociology, LSE)
This course offers an introduction to MCA, which is a method that allows researchers to observe the patterning of complex data sets through representing categorical variables as points in N-dimensional space. Although it was developed from the later 1960s, MCA has not previously had a large Anglophone following, but it is an increasingly popular method because of (a) its association with Pierre Bourdieu’s high profile sociology, (b) its capacity to lend itself to visualisation of clusters and (c) its potential for mixed methods research.
This course is suitable for:
PhD students, post-doctoral fellows and academic staff in the social sciences, interested in one of the main methods for the clustering of categorical data
those interested in learning about the methods used by Pierre Bourdieu for the analysis of cultural fields and social relations
market researchers, other commercial researchers, and public sector professionals wishing to learn MCA as a means of clustering complex data sets, and presenting attractive and intuitive visualisations.
This course will provide students with:
a comprehensive introduction to MCA
training in how to use SPAD software
awareness of key exemplars in social science using MCA and an awareness of the theoretical principles it draws upon.
No statistical knowledge is necessary, but it will be advantageous. Applicants must be at PhD level or higher.
This course will offer a comprehensive introduction to the principles of multiple correspondence analysis. Comprehensive training is also provided in using SPAD software, the most accessible and flexible package to use when carrying out MCA.
Issues covered include mathematical principles of geometric data analysis, the difference between the active space of modalities and the use of supplementary variables, coding issues, working with the cloud of modalities and the cloud of individuals, clustering methods within MCA, and the use of inferential statistics within MCA.
The course is designed to allow the beginner to grasp basic mathematical principles of geometric data analysis. The course will be delivered by a series of lectures by leading international experts in MCA in the morning, with practical sessions in a computer lab in the afternoons.
The following teaching schedule is indicative only, and is subject to change.
Introduction to MCA and SPAD
Creating and refining active spaces and using supplementary variables
Clustering within SPAD
Inferential statistics and MCA
Exemplars and MCA
Assessment is in the form of a practical assignment completed over the week of the course.
Brigitte Le Roux and Henry Rouanet (2010) Multiple Correspondence Analysis. QASS nº 163. SAGE.