Lani Freeborn, University of Amsterdam

Network analysis is a useful, exploratory tool for identifying patterns in multivariate data (Bringmann et al., 2016). With psychological networks, variables are modelled as nodes in a system, and edges between nodes represent conditional pairwise associations. In second language acquisition research for example, network analysis has been used to examine the interplay between motivation and self-regulation (Rahimi et al., 2025) and task engagement and emotions (Liang et al., 2025). While network analysis is closely related to factor analysis, they each reflect different theories (Fried, 2020). Factor-based approaches assume that variables correlate because they measure the same underlying construct (i.e., a latent variable). In contrast, network approaches assume direct causal relationships between variables, whereby it is the network of relationships between variables that constitutes the construct. From this perspective, network approaches are considered to have a close, conceptual alignment with Complex Dynamic Systems theory (CDST). In this presentation, I will introduce the theoretical foundations of network analysis, highlighting its alignment with CDST. I will demonstrate how network analysis can be utilised in the field of second language acquisition, viewing individual differences and language development from a systems perspective. The presentation will include empirical examples of network models with both cross-sectional and longitudinal data.

Part 1 in a series of talks on quantitative data analysis in applied linguistics research