A didactic model for developing sociolinguistic appropriateness in foreign language academic writing for linguistics undergraduates through the integration of prompts based on artificial intelligence and peer assessment
A didactic model for developing sociolinguistic appropriateness in foreign language academic writing for linguistics undergraduates through the integration of prompts based on artificial intelligence and peer assessment
I. E. Rodikov
TLDR
The proposed model is a promising didactic solution for organizing step-by-step L2 academic writing within face-to-face educational settings and is proposed as a reproducible formative assessment procedure for L2 academic writing by integrating prompts based on artificial intelligence and peer assessment.
Abstract
The research aims to theoretically substantiate and describe a didactic model for developing sociolinguistic appropriateness in foreign language (L2) academic writing among linguistics undergraduates by integrating prompts based on artificial intelligence (AI-based prompts) and peer assessment. The article presents the model’s structure and details its implementation within a face-to-face English academic writing course as part of the “Practical Course of the First Foreign Language” discipline. The described model involves a sequential workflow: drafting, obtaining AI-based prompts via standardized requests (considering genre and addressee), self-correction, peer review by two classmates using an analytical rubric, and final editing. The analytical rubric provides a unified criterion-based foundation for the formative assessment of the text’s sociolinguistic appropriateness and is utilized both during peer review and when evaluating AI-generated recommendations. The scientific novelty of the study lies in the development of a reproducible formative assessment procedure for L2 academic writing, in which AI-based prompts and peer assessment are unified by a common criteria framework. As a result of testing the model through a formative experiment, it was established that the score increase in the experimental group significantly outperformed the control group: 3.6 points on the total scale and 1.2 points for the sociolinguistic appropriateness component, compared to 1.4 and 0.4 points, respectively. Approximately 40% of the edits in the experimental group were related to register, speech etiquette, audience awareness, and the degree of hedging (categoricalness). The findings suggest that the proposed model is a promising didactic solution for organizing step-by-step L2 academic writing within face-to-face educational settings.