VIETTESOL INTERNATIONAL CONVENTION 2025

There has been a switch from lesson content mastering to skills developing as the movement of competency in the 21st century. To respond to this trend in practical requirements, University of Economics, Ho Chi Minh City (UEH) has focused on providing English for Specific Purposes (ESP) courses where students expect to learn knowledge and skills relating to their fields of study. These courses include four key majors at UEH and encompass four basic skills in English language learning. Various prior studies have been conducted for seeking ways to assess and evaluate English written compositions. Among these, building and validating appropriate scoring scales have yielded positive results in reducing human raters’ burden, especially in ESP writing assessment including many unique writing tasks. Therefore, ESP teachers can have more chances to give feedback and design remedial lessons accordingly for their students. With the advent of Artificial Intelligence (AI) technology, teachers are using numerous AI tools as personal assistants for designing their own scoring scales. This case study used the Claude 3.5 Sonnet for creating a Python tool tailored to compare 224 Law students’ written compositions with a given sample. This newly created tool was then validated by comparing its automated scores with human ratings and produced a positive correlation. Therefore, this tool can reflect human ratings and have the potency to reduce teacher scoring workload. Finally, pedagogical implications were discussed relating to this tool use and how to deal with some frequent errors found in students’ compositions.