Functional Categorical Data Analysis of SJR Rankings

W. R. Pérez Rocano, A. G. López Herrera, M. Escabias Machuca

The analysis of the temporal dynamics of bibliometric indicators such as the SJR index requires methodologies capable of modeling both their temporal evolution and categorical nature. In this work, the Categorical Functional Data Analysis framework is applied to study the dynamics of scientific journals classified into quartiles over time. Each journal is modeled as a categorical functional trajectory, where the states represent SJR quartiles and their evolution describes patterns of persistence and transitions between quartiles.

From these trajectories, principal components are estimated using optimal encoding functions, which are obtained by solving an eigenvalue problem associated with the covariance structure of the process.

The results show that a reduced number of components explains a significant proportion of variability, allowing the identification of patterns of stability, mobility, and consolidation in journal rankings. This approach provides a robust tool for analyzing longitudinal categorical data, extending classical functional data analysis methods to discrete settings.

Keywords: Categorical Functional Data, Functional Data Analysis, Functional Principal Component, SJR

Functional Data Analysis, Spatial and Spatio-Temporal Statistics
September 2, 2026  5:40 PM
Aula 26

N. A. Cruz Gutierrez, A. Mir, I. Garcia-Mosquera, A. Tobar-Nicolau, M. Miró Julià, R. Alberich

N. Lopes Garcia, V. Freguglia

F. J. Soto