Human-Machine Collaboration in Emotionally Responsive Teaching Practice

Abstract

Emotional analytics tools are increasingly embedded within educational platforms, yet their pedagogical value depends not on the data they generate but on the relational conditions in which that data is interpreted. This study examined relationships among seven dimensions of socio-emotional experience: emotional awareness, relational belonging, engagement, purpose, teaching climate, SEL strategy use, and perceived usefulness of Microsoft Reflect, in a university English-medium instruction context in Japan (N = 89), following a three-week classroom implementation. Using a quantitative correlational design and a SEL–Reflect survey, results revealed that Teaching Climate demonstrated the strongest pattern of positive associations across all subscales, while Emotional Awareness alone showed no significant relationship with Reflect Usefulness. These findings suggest that emotional self-awareness does not, by itself, make emotional analytics educationally valuable; rather, their value depends on the quality of classroom relationships and teachers' capacity to interpret affective data within its relational and pedagogical context. This has particular relevance for students managing anxiety or social-emotional difficulties, for whom a responsive teaching climate may be a precondition for meaningful engagement with emotional self-report tools. Positioning teachers as central agents within human-machine hybrid teams, we argue that emotional analytics function most effectively as cognitive and relational scaffolds that extend professional judgment rather than automate educational decision-making. The study contributes to discussions of teacher agency, emotionally responsive pedagogy, and the ethical integration of emotional analytics in higher education.