2025 ASEE Annual Conference & Exposition

Engineering curricula in higher education prioritize the development of students’ technical knowledge and skills to a certain extent. The cultivation of well-balanced and structured writing skills, essential for knowledge construction, is often overlooked. Students have used generative AI tools for their writing tasks in their courses. However, excessive dependence on generative AI for writing tasks can negatively impact engineering education. One strategy for supporting knowledge construction and writing skills can involve text visualization while students are writing essays or reports in their courses. Based on the strategy, this study suggests a more thoughtful use of AI to assist student writing, offering an alternative to the careless adoption of generative AI. Specifically, the goal of the project is to explore an approach to improving the knowledge construction process with technology-enabled writing projects in undergraduate engineering courses. The research team has developed a knowledge visualization tool, known as KVIS (Knowledge Visualization Intelligent System), designed to visualize the structure of a students’ writing. KVIS illustrates the overall structure, relationships between keywords, and connections among concepts. This enables students to better comprehend their current writing and effectively reconstruct ideas and opinions based on their understanding. The team (1) assessed the effects of a knowledge visualization tool in writing projects on undergraduate engineering students’ knowledge construction outcomes and (2) examined the potential prediction power of self-regulated learning and course grades on writing scores. First, a one-way repeated measures ANOVA (Analysis of Variance) was conducted to examine changes in writing performance over time. It was found that engineering students’ writing performance statistically significantly increased over time. Second, regression analysis was conducted to investigate the prediction power of self-regulated learning and course grades on students’ writing performance. The results showed that self-regulated learning and course grades did not significantly predict writing performance. This study demonstrated possibilities to provide engineering students with writing skill development opportunities through text visualization support.