Survey responses on self-regulation, rhetorical awareness, writing motivation and generative AI use among master's students (n=450)

Published: 20 November 2026| Version 1 | DOI: 10.17632/sn2tx7bthh.1
Contributor:
Hamza Miftah

Description

This dataset supports an exploratory mixed-methods study of AI-mediated academic writing among French graduate students. The study examined how quantitative profiles of generative AI use and writing-related characteristics (writing self-regulation, rhetorical awareness, and writing motivation) relate to graduate students' situated academic writing practices. It focused on the possibility that different configurations of these dimensions could relate to distinct writing practices and purposes. The dataset contains responses to an online survey administered in French between January and May 2026 to master's students from 21 programs at a single university. Participation was voluntary and anonymous. The dataset includes the 450 complete responses; incomplete responses were excluded. Latent profile analysis was used to identify a three-profile solution: limited AI use (n = 179), frequent AI use with lower scores on the other three dimensions (n = 121), and frequent AI use with higher scores on those dimensions (n = 150). AI-use frequency differentiated the profiles most, and frequent AI use did not systematically coincide with weaker writing-related engagement. The profiles identified are descriptive rather than fixed. The survey data were complemented by qualitative case studies based on interviews with students, their academic texts, and their AI interaction records, which were used to interpret the quantitative profile tendencies in more concrete terms.

Files

Categories

Motivation, Academic Writing in English, Academic Writing, Self-Regulation, Writing, Generative Artificial Intelligence

Licence