How LLM Could Be Used for Generating Text to Facilitate Multimedia-Based Physics Assessment?

6 Sept 2025, 11:20
20m
Room 0.81 (ELTE TTK)

Room 0.81

ELTE TTK

Oral presentation AI in Multimedia-Based Physics Assessment and Feedback Oral Presentations

Speaker

Prof. Tomasz Greczyło (Institute of Experimental Physics, University of Wrocław)

Description

The use of physics textbooks in education and their impact on national teaching and learning practices has a long-standing tradition and has been investigated from various perspectives [1, 2, 3]. Recently, there has been growing interest in using literary texts in science education. One particular approach to structuring assessments using different types of texts has been presented in a nationally published primary school physics workbook [4] associated with the physics textbook.

This study aims to highlight the results of research in which various freely available Large Language Models (LLMs) were used to analyse literary texts originally written in Polish. The goal was to identify physics-related topics and phenomena that could be assessed using these texts, based on the author's initial judgment. Following this analysis, the LLMs were tasked with generating similar texts that could potentially hold educational value for teaching and learning physics. The outcomes of this work are intended to be presented and discussed from multiple educational perspectives.

Contribution categories - primary focus University
Contribution categories - type Research oriented

Author

Prof. Tomasz Greczyło (Institute of Experimental Physics, University of Wrocław)

Presentation materials