ISE PhD course: Applied Multiple Correspondence Analysis (MCA) for the social and human sciences
Doctoral School of Social Sciences and Business at Roskilde University
Multiple Correspondence Analysis (MCA) is a geometric data analysis technique designed to uncover patterns and structures in complex categorical datasets. Rooted in the tradition of French data analysis pioneered in the 1960s by French statistician Jean-Paul Benzécri, MCA represents data as clouds of points in a multidimensional space. The position and distance between these points can tell us about the relationships within the data. This way MCA allows researchers to visualize associations between multiple categorical variables in a low-dimensional space. This in turn makes MCA particularly well-suited for analyzing the relational structures of survey- as well as prosopographic data, and for developing typologies, and classifications—making it a powerful tool for social scientists working with empirical data.
Within the social and human sciences, MCA gained prominence through the work of Pierre Bourdieu, who used it extensively to map social space and different fields and analyze the distribution and composition of different forms of capital (cultural, economic, social, symbolic and in special cases academic capital) among different social groups. In seminal works such as Distinction (1984), Homo Academicus (1988), State Nobility (1996) and The Social Structures of the Economy (2005), Bourdieu employed MCA to empirically ground his theoretical concepts, demonstrating how social positions and lifestyles could be visualized and interpreted through relational data structures. His use of MCA not only spearheaded the use of the method but also showcased its potential for bridging quantitative analysis with rich sociological theory.
Students will learn how to prepare data, conduct MCA using statistical software (e.g., R), interpret graphical outputs, and critically reflect on the epistemological implications of the method. The course also explores how MCA can be used to operationalize concepts such as capital, field, social position and lifestyle, and how it supports mixed-methods research designs. The course emphasizes hands-on experience with MCA software (e.g., R) and the practical application of MCA to the phd. Students own data.