Från AI prompt till betyg

: En deltagande observation av bedömningsnätverk i gymnasieskolan

Abstract

This study explores how upper secondary school teachers use generative artificial intelligence, with a focus on ChatGPT, in their assessment practices. Inspired by actor-network theory, the study was conducted through participant observation, following two teachers in their interactions with ChatGPT. By mapping the actors in the network, their influence and impact on assessment practices are analyzed. The results indicate that ChatGPT can contribute to optimizing and supporting teachers’ grading and feedback processes, while also raising questions about transparency, control, and potential shifts in responsibility. Using an actor-network perspective, the study identifies ChatGPT alongside other actors, not merely as a tool, but as an actor that reshapes assessment practices.

Date of Award	2025-Jun
Original language	Swedish
Supervisor	Lars-Erik Nilsson (Supervisor), Anders Eklöf (Assessor) & Andreas Redfors (Examiner)