Search

Abstract Background Self-regulated learning (SRL) strategies can be domain specific. However, it remains unclear whether this specificity extends to different subtopics within a single subject domain. In this study, we collected data from 210 college students engaged in a computer-based learning environment to examine the heterogeneous manifestations of learning behaviors across four distinct subtopics in introductory statistics. Further, we explore how the time spent engaging in metacognitive strategies correlated with learning gain in those subtopics. Results By employing two different analytical approaches that combine data-driven learning analytics (i.e., sequential pattern mining in this case), and theory-informed methods (i.e., coherence analysis), we discovered significant variability in the frequency of learning patterns that are potentially associated with SRL-relevant strategies across four subtopics. In a subtopic related to calculations, engagement in coherent quizzes (i.e., a type of metacognitive strategy) was found to be significantly less related to learning gains compared to other subtopics. Additionally, we found that students with different levels of prior knowledge and learning gains demonstrated varying degrees of engagement in learning patterns in an SRL context. Conclusion The findings imply that the use—and the effectiveness—of learning patterns that are potentially associated with SRL-relevant strategies varies not only across contexts and domains, but even across different subtopics within a single subject. This underscores the importance of personalized, context-aware SRL training interventions in computer-based learning environments, which could significantly enhance learning outcomes by addressing the heterogeneous relationships between SRL activities and outcomes. Further, we suggest theoretical implications of subtopic-specific heterogeneity within the context of various SRL models. Understanding SRL heterogeneity enhances these theories, offering more nuanced insights into learners’ metacognitive strategies across different subtopics.

Educational data mining has allowed for large improvements in educational outcomes and understanding of educational processes. However, there remains a constant tension between educational data mining advances and protecting student privacy while using educational datasets. Publicly available datasets have facilitated numerous research projects while striving to preserve student privacy via strict anonymization protocols (e.g., k-anonymity); however, little is known about the relationship between anonymization and utility of educational datasets for downstream educational data mining tasks, nor how anonymization processes might be improved for such tasks. We provide a framework for strictly anonymizing educational datasets with a focus on improving downstream performance in common tasks such as student outcome prediction. We evaluate our anonymization framework on five diverse educational datasets with machine learning-based downstream task examples to demonstrate both the effect of anonymization and our means to improve it. Our method improves downstream machine learning accuracy versus baseline data anonymization by 30.59%, on average, by guiding the anonymization process toward strategies that anonymize the least important information while leaving the most valuable information intact.

In this paper, we propose a new method for selecting cases for in situ, immediate interview research: detector-driven classroom interviewing (DDCI). Published work in educational data mining and learning analytics has yielded highly scalable measures that can detect key aspects of student interaction with computer-based learning in close to real-time. These measures detect a variety of constructs and make it possible to increase the precision and time-efficiency of this form of research. We review four examples that show how the method can be used to study why students become frustrated and how they respond, how anxiety influences how students respond to frustration, how metacognition interacts with affect, and how to improve the design of an adaptive learning system. Lastly, we compare DDCI to other mixed-methods approaches and outline opportunities for detector-driven classroom interviewing in research and practice, including research opportunities, design improvement opportunities, and pedagogical opportunities for teachers.

Metacognition is a valuable tool for learning, due to its role in self-regulated learning. However, online learning settings bring new challenges for engaging in metacognition​​ given the unique opportunities and challenges presented by the online space, especially for diverse populations and students underrepresented in STEM (UR-STEM). Thus, we investigated whether a relation existed between college STEM students’ metacognition—measured by their spontaneously produced metacognitive phrases in online course discussions forums—and their success in an online STEM college course—measured by their final course grade. Using Bayesian generalized linear models, we examined whether this relation differed for UR-STEM compared to non-UR-STEM students and whether related course behaviors (i.e., engagement and verbosity) and prior knowledge predicted variance in course grade. Metacognition plausibly predicted course grade and we found no plausible differences between UR- and non-UR-STEM students, suggesting that the online space could afford students from diverse groups the capacity to engage equally in a critical aspect of self-regulated learning: metacognition. Implications of the results for teaching and learning STEM content in the online space are discussed.