Improving Students' Understanding through Metacognition About Instructor Feedback: A Causal Modeling Approach

Previous research has documented learning benefits from metacognitive engagement and suggested many instructional practices to prompt metacognition in students. For example, a scaffolded curriculum that includes effective feedback may enable students to construct durable and nuanced knowledge. However, as ethical and logistical constraints prevent randomly assigning students to conditions or offering treatments in isolation from other confounding variables, true experiments are challenging to conduct in the scholarship of teaching and learning (SoTL) and identifying causal relationships between teaching strategies and metacognition or metacognition and learning remains elusive. In this study, using an interrupted time series design and a series of structural equation models, we demonstrate the potential of causal modeling to analyze pedagogical data and evaluate a novel metacognitive intervention. A total of 445 undergraduate students in a psychological science research methods course at a large public university in the United States of America in 2023 were given the option to submit exam corrections based on feedback and formatted as an assignment designed to prompt metacognitive engagement. Exam performance over the academic term was compared between those who completed the intervention and those who opted out. Exam performance for those students who opted in was improved beyond what was predicted by their baseline trajectory. Through a series of eight nested structural equation models, this study presents evidence that the intervention was causally related to performance.