Back to Search Start Over

USING LARGE LANGUAGE MODELS FOR ACADEMIC WRITING INSTRUCTION: CONCEPTUAL DESIGN AND EVALUATION OF THE SOCRAT PROJECT.

Authors :
Spirgi, Lukas
Seufert, Sabine
Source :
Proceedings of the IADIS International Conference on Cognition & Exploratory Learning in Digital Age; 2024, p3-10, 8p
Publication Year :
2024

Abstract

Academic writing has undergone significant evolution due to advancements in AI. Students are leveraging AI in diverse ways for their studies. This study introduces a course design (SOCRAT) to teach students genre-based academic writing through AI. Genre-based academic writing is an educational strategy instructing students in the writing techniques and norms pertinent to their specific academic disciplines. AI is utilised as a personal training system and research assistant in this proposed course design. Students require cognitive and metacognitive knowledge to effectively work with AI tools. The SOCRAT design is based on the concept of mastery learning to ensure that students build their competencies. An initial evaluation of the prompts developed for this design indicates that LLM can particularly assist students in analysing their written text and providing suggestions for enhancement. This can help students develop their genre-based writing skills. The models are not yet convincing for other types of tasks where the LLMs are required to give exact answers. [ABSTRACT FROM AUTHOR]

Details

Language :
English
Database :
Supplemental Index
Journal :
Proceedings of the IADIS International Conference on Cognition & Exploratory Learning in Digital Age
Publication Type :
Conference
Accession number :
181126526