Supporting pre-service teachers’ diagnostic skills: Analysing judgment accuracy and chatbot impact in a biology classroom simulation

Background:

Novice teachers often struggle to apply their content and pedagogical content knowledge in real teaching situations, a phenomenon known as the theory–practice gap. Classroom simulations offer an opportunity to bridge this gap by integrating practical, risk-free experiences into early university teacher education. Enriching these simulations with chatbots could enhance this process by providing tailored, adaptive learning support, potentially improving pre-service teachers' diagnostic skills.

Objectives:

This study investigates pre-service biology teachers' ability to diagnose virtual students' evolutionary explanations in the classroom simulation SCRBio and the effect of a retrieval-based chatbot system.

Methods:

A sample of 107 pre-service biology teachers diagnosed 3153 virtual students' evolutionary explanations using three SCRBio settings: without a chatbot (SCR-only), with a knowledge bot (SCR-Kbot), and with both a knowledge bot and a process-based feedback bot (SCR-Kbot+Pbot).

Results and Conclusions:

Pre-service biology teachers were able to broadly categorize student explanations in a classroom simulation; however, they struggled to accurately diagnose specific misconceptions, especially in mixed-scientific responses. Regression analysis revealed that pedagogical content knowledge and cognitive load had a significant influence on judgement accuracy. Although chatbot-based support was accessed more frequently when combined with process prompts (SCR-Kbot+Pbot group), it did not result in significant improvements in accuracy. These findings highlight the need for targeted training for diagnostic skills and suggest that the integration of adaptive digital tools alone may not be sufficient to enhance diagnostic competence without deeper pedagogical embedding.