Instantiation of the DiaCoM framework highlighting multimodal conceptualization and measurement of teacher noticing in human-AI-partnered classrooms.

Instantiation of the DiaCoM framework to illustrate the focus of the current empirical study, involving teachers’ diagnostic behaviors, personal characteristics, and situational cues (adapted from Loibl et al., 2020).
Investigating analytics-supported teacher noticing using fine-grained data on student-AI and teacher-AI interaction. As represented in the Figure, we study the interplay between situational cues (i.e., students’ in-the-moment struggle when learning the AI system), teachers’ personal characteristics (i.e., prior knowledge of low versus high-performing students), and teachers’ diagnostic behaviors. With the availability of novel data streams to operationalize teacher noticing in the wild, we can also generate new insights into what effects teacher noticing has on student learning.