A two-staged SEM-ANN approach to predict learning presence in online foreign language education: The role of teaching presence and online interaction

Abstract

While the theoretical value of the Community of Inquiry (CoI) framework in comprehending online learning experiences has been acknowledged, the newly introduced CoI element—learning presence—has received insufficient attention in the field of foreign language (FL) education. Drawing on the constructivist stimulus-mediation-response approach, this study investigated the predicting effects of teaching presence and online interaction on learning presence. Data were collected from 460 college-level online Chinese as a Foreign Language (CFL) learners at seven Chinese higher education institutions. Partial least squares structural equation modeling (PLS-SEM) was used to explore the linear relationships, followed by the artificial neural network (ANN) technique to assess the relative importance of predictors based on the nonlinear relationships between variables in the research model. The results support most of the predictive effects of teaching presence and online interaction on learning presence variables (self-efficacy, metacognitive self-regulation, and metacognitive co-regulation). Teaching presence, learner-instructor interaction, and learner-learner interaction were the most influential predictors of self-efficacy, metacognitive self-regulation, and metacognitive co-regulation, respectively. Furthermore, the mediating role of online interaction between teaching presence and learning presence was partially supported. The findings highlight the critical roles of teaching presence and online interaction in fostering online FL learners' active and responsible learning while offering valuable insights into the design of online language courses.