Supporting data for "examining teachers’ dialogic strategies and cognitive processes when interacting with AI simulated students"

This dataset was compiled by Wu, Jiajun (Faculty of Education, HKU) to support his PhD research on teacher noticing within AI-enhanced educational environments. It comprises data from 42 chemistry teachers who engaged in teaching sessions with AI-simulated students. The dataset includes four primary components:Transcripts (<code>Transcript.zip</code>): Contains 49 session folders, each holding a classroom dialogue transcript (<code>Dialogue.docx</code>) and a post-class reflection report (<code>Report.docx</code>). Sessions cover specific chemistry topics such as combustion principles and reaction rates.Coded Data (<code>Coding.xlsx</code>): Features 119 coded dialogue excerpts extracted from the dialogues and reports. It includes 27 variables detailing teacher noticing processes across three phases: Identification (e.g., recognizing student misconceptions), Interpretation (e.g., analyzing knowledge components like PCK), and Decision (e.g., selecting adaptive pedagogical moves or scaffolding prompts). It also codes for reflective qualities in post-class reports.Survey Instruments &amp; Data (<code>Questionnaires.docx</code> &amp; <code>.xlsx</code>): Provides pre- and post-task survey responses from the 42 participating teachers. The data includes demographic information (teaching experience, rank) and 31 Likert-scale items measuring beliefs about classroom discussions, plus 8 items comparing AI-simulated experiences to real classrooms, alongside open-ended qualitative feedback.This resource is designed for researchers investigating teacher cognition, AI simulation efficacy, and pedagogical content knowledge (PCK) development, offering a linked structure between raw interaction logs, fine-grained behavioral coding, and self-reported teacher perceptions.