Using Trace Data of Secondary Students to Understand Metacognitive Processes in Writing From Multiple Sources

Background: Integrating information from multiple sources is a common yet challenging learning task for secondary school students. Many underuse metacognitive skills, such as monitoring and control, which are essential for promoting engagement and effective learning outcomes. Objective: This study aims to examine the relationship between metacognitive processes and the quality of writing from multiple sources with diverse language backgrounds. Methods: To understand these processes, we conducted a laboratory study with 162 secondary students from diverse language backgrounds (English, German and Finnish). We collected trace data about metacognition while students were reading and writing in a digital learning platform. These data, along with the language used by students to produce their writing, were analysed to determine the association of metacognition with essay scores obtained using both automated and human evaluations. Result and Conclusion: Our findings indicate that students from different language backgrounds exhibit varying performance levels detectable by automated scoring. Secondary school students showed limited metacognitive processes in multi-source writing; this contrasts with findings from previous studies conducted in higher education. These findings can inform the development of analytics-based tools to support secondary students' writing through trace data and automated essay scoring, and provide evidence of the need for targeted interventions to assist and support secondary school students in improving their writing from multiple sources.