Key points are not available for this paper at this time.
Abstract The rapid adoption of generative artificial intelligence (genAI) applications, such as ChatGPT, underscores their potential to significantly enhance writing. However, genAI efficacy varies among users and use cases, likely due to insufficient understanding of their operations and limitations. This study investigates the effect of genAI and targeted instruction on writing productivity and quality among graduate students on graded professional writing assignments. Employing a mixed-methods design, we compared students’ performance on professional writing tasks conducted without and with genAI assistance. Comprehensive genAI instruction was provided prior to tool usage. Results showed genAI, with proper instruction, reduced writing time by 56.7% and improved writing quality from an A- to an A on average. Productivity benefits were observed across native English (NES) and non-native English speaking (NNES) students. Quality benefits were more prominent in the NNES students, suggesting a potential leveling effect of genAI in support of writing in a second language. This study highlights the importance of tailored educational strategies in leveraging genAI tools to build students’ competency to maximize benefits and mitigate limitations.
Pensky et al. (2025) studied this question.