Implicações da Correção Automatizada de Redações para a Agência do Interator: um estudo da plataforma coRedação

Abstract

This study aims to critically analyze the free version of the coRedação platform, which utilizes artificial intelligence to assess students' essays, investigating whether its use promotes the agency of the interactor and enhances writing skills development. To promote the discussion on automated correction, we analyzed the platform using the Close Reading methodology, submitting four distinct essays totaling ten submissions to observe its performance regarding various textual qualities. The analysis revealed that, although the platform is effective in verifying adherence to the argumentative essay genre and assigning grades, it does not necessarily promote student agency. This is due to the lack of personalized feedback, which fails to encourage a conscious construction of meaning, prioritizing a textual production adapted to standardized formats. Furthermore, the difficulty in tracking progress through revisions may weaken the sense of autonomy, as students’ contributions may not have a tangible impact on their final scores. This research stands out for its relevance in addressing a significant gap in the field. Most existing investigations are predominantly theoretical, limiting practical understanding of the functionalities of automated correction artifacts. By analyzing these tools in a real educational context, we aim to evaluate their applicability in educational environments, considering the increasing implementation of technologies in education.