Tracing Causal Relationships in L2 Learning via Granger Causality.

Abstract

In this article we explore the use of Granger Causality as a method to shed light on causal relationships in studies on (second) language acquisition. Understanding such relationships remains a central challenge in longitudinal SLA research, where both learner variables and measures of linguistic development interact and evolve dynamically over time. We first explain the concept of Granger Causality and show how the method works, before providing an illustration using weekly measures of L2 development (C-Test performance and speech fluency) and individual difference variables (operation span, motivation and wellbeing) to explore their possible causal relationships in a longitudinal study with later-life language learners. The results reveal several asymmetric patterns. The findings are discussed in relation to the existing literature on later-life language learning and demonstrate that Granger causality is a valuable method for examining the directionality of relationships between variables in longitudinal data.