Baltimore , Maryland
June 25, 2023
June 25, 2023
June 28, 2023
Diversity and NSF Grantees Poster Session
In the last two decades, flipped learning has become one of the pedagogies to integrate active learning in a classroom. Although flipped learning has tangible benefits for learning and engagement, student resistance remains a challenging issue. This struggle is most evident in pre-class learning, which leads to inadequate preparation for the in-class engagement exercises, including answering conceptual questions and solving procedural problems.
In a course in Numerical Methods taught in the department of mechanical engineering at a large southeastern university, we alleviated this resistance to pre-class preparation via adaptive learning platform (ALP) lessons. ALPs use machine learning algorithms to deliver personalized learning activities and valuable feedback to students reliably.
One of the benefits of using an ALP is the significant amount of data it collects about student behavior and engagement with the course material. Since these ALP lessons were used from the beginning of the semester onward, we proactively identified low-performing students who may have needed support to improve their performance in the course. Based on the results of a descriptive analysis performed on ALP collected during Fall 2021 and Spring 2022, we found that students who were most likely to exhibit low performance (as defined by a course grade of C or below) did not complete their ALP lessons on time, scored lower on the ALP assessments, made a higher number of attempts, and spent less time on the instructional content versus the assessment questions with the platform.
In Fall 2022, we used decision tree models with the ALP data to identify potentially low-performing students in the academic semester's second, third, and fourth weeks. Based on this, we extended an official invitation for one-on-one support and advising to 18 students out of a class of 62 who were identified as potentially low performing during the first four weeks. Six of them accepted the invitation and met regularly at a scheduled time with the instructor or the teaching assistants. In this paper, we will discuss the performance of all 18 students and compare those who accepted versus declined the invitation. Considering that the decision tree models may not have identified some low-performing students, necessary adjustments to the model will be made for better identification in Spring 2023, with subsequent results available at the time of the conference.
Kaw, A., & Yalcin, A., & Clark, R. M. (2023, June), Board 422: Using Adaptive Learning Platform Metrics for Early Identification and Personalized Support of Low-Performing Students Paper presented at 2023 ASEE Annual Conference & Exposition, Baltimore , Maryland. 10.18260/1-2--42761
ASEE holds the copyright on this document. It may be read by the public free of charge. Authors may archive their work on personal websites or in institutional repositories with the following citation: © 2023 American Society for Engineering Education. Other scholars may excerpt or quote from these materials with the same citation. When excerpting or quoting from Conference Proceedings, authors should, in addition to noting the ASEE copyright, list all the original authors and their institutions and name the host city of the conference. - Last updated April 1, 2015