Project title
Logistic multidimensional data analysis
Research question
How can we analyse and visualise complex data sets with multiple categorical variables?
Project description
Multivariate data are often analyzed using techniques like principal component analysis and multidimensional unfolding. Both methods require numeric data, meaning the variables should be measured on an interval or ratio scale. However, in the social and behavioural sciences, measurements are often categorical, such as binary, ordinal, or nominal. For categorical data, logistic models are the most useful.
Over the past two decades, Mark de Rooij has been developing logistic multidimensional data analysis techniques. In this project, he plans to write a book and create accompanying software to make these tools accessible to a wider scientific audience.
Selected publications
- De Rooij, M., Busing, F. (2024). Multinomial Restricted Unfolding. Journal of Classification, 41, 190–213. https://doi.org/10.1007/s00357-024-09465-3
- De Rooij, M. (2024) A new algorithm and a discussion about visualization for logistic reduced rank regression. Behaviormetrika, 51, 389–410. https://doi.org/10.1007/s41237-023-00204-3
- De Rooij, M., Breemer, L., Woestenburg, D., and Busing, F. (2023, submitted). Logistic Multidimensional Data Analysis for Ordinal Response Variables using a Cumulative Link function. Preprint available: https://doi.org/10.48550/arXiv.2402.07629