Collaborative Research: EAGER: Developing and Optimizing Reflection-Informed STEM Learning and Instruction by Integrating Learning Technologies with Natural Language Processing

This project aims to enhance student learning and engagement in large lecture STEM courses by developing, optimizing, and evaluating a digital learning environment called CourseMIRROR. CourseMIRROR uses Natural Language Processing (NLP) algorithms and techniques to prompt and scaffold students to create in-depth reflections on their learning experiences. By closely working with a socially and culturally diverse group of students and instructors in public universities and community colleges, the project will directly affect hundreds of students through evidence-based pedagogies and the way educators provide opportunities for learning and engagement. Since we purposefully selected to work with diverse students across institutions, findings will be generalizable to the college student population. Also, the multidisciplinary nature of the project team and work ensures that our results will be reached across traditional disciplinary silos, generating impact in multiple fields, including NLP, Artificial Intelligence (AI), Human-Computer Interaction (HCI), learning sciences, and STEM education. By examining students’ learning through purposeful reflection and feedback loops, this work has the potential to provide a route to personalized learning with innovative approaches to problems vital in the increasingly global economy, thereby opening an important new direction of research in learning sciences and emerging technologies. The proposed project will explore the role of the reflection-informed learning and instruction (RILI) model on students’ engagement and learning outcomes in large lecture STEM courses. The research team will develop and optimize the CourseMIRROR digital learning system that leverages NLP techniques to prompt and scaffold students to write detailed reflections and generate reflection summaries for each lecture. Specifically, this project will incorporate three lines of research: 1) the role of the RILI model on students’ motivation, emotions, and learning, 2) the effectiveness of NLP in creating personalized learning experiences, summarizing reflections in a meaningful way, and evaluating the quality of reflections, and 3) value and design of digital learning tools to improve students’ engagement and learning. This project leverages NLP and HCI techniques and connects them with the RILI model. The aim of combining these approaches emerges to support the innovative and unconventional approach to research, pedagogical strategies, and improved student outcomes. How students learn through iterative cycles of critical reflection and how to effectively utilize and optimize prompts and feedback is not yet well understood or studied. Equally important is how instructors use the process of reflective practice to inform and transform instruction. This project is novel in this respect, as researchers have yet to conduct studies in which these questions are jointly explored and help us explore how learning and engagement can be enabled, improved, and supported across different classes using digital tools, social interactions, and practices. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2623779 | Program: 01002324DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Muhsin Menekse | Institution: Southern Methodist University, DALLAS, TX | Award Amount: $26,573 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2623779 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2623779.html