Educational data mining (EDM) is an emerging interdisciplinary field that utilizes a machine learning (ML) algorithm to collect and analyze educational data, aiming to better predict students' performance and retention. In this WIP paper, we report our methodology and preliminary results from utilizing a ML program to assess students’ motivation through their upper-division years in the XYZ project-based learning (PBL) program. ML, or more specifically, the clustering algorithm, opens the door to processing large amounts of student-written artifacts, such as reflection journals, project reports, and written assignments, and then identifies keywords that signal their levels of motivation (i.e., extrinsic vs. intrinsic). These results will be compared against other measures of motivation, including student self-report, faculty observation, and externally validated surveys. As part of a longer-term study, this pilot work sheds light on the key question for student success and retention: how does student motivation evolve through the 3rd and 4th years in college?
The purpose of this research project is to gain insights into learners’ motivation levels and how it evolves during the last two years in college, as well as to extend current Educational Data Mining research and Machine Learning analysis described in the literature. It is significant on two fronts: 1) we will extend the ability of ML in analyzing reflective written artifacts to explore student physiological and emotional development; 2) the longitudinal study will help monitor the progressive change of motivation in college students in a PBL environment.
Preliminary results from an initial pilot study are promising. By analyzing written reflection journal entries from previous students, the ML algorithm has differentiated keywords into three student motivation levels: “high”, “neutral” and “low”. Using supervised classes, for example, the ML algorithm differentiated words in the highly motivated student text such as “team” and “learning”, while the text coded as low motivation included “use”, “pushed” and “nothing”.
For our future research, we aim to create a dictionary that identifies words/phrases related to positive/negative motivation. We will extend the pilot study to a longitudinal evaluation of student motivation over four semesters of engineering education as well as prediction of student success in a PBL environment.
2022 ASEE Annual Conference & Exposition
Pluskwik, E., & Wang, Y., & Singelmann, L. (2022, August), Work-in-Progress: Understanding learners' motivation through machine learning analysis on reflection writing. Paper presented at 2022 ASEE Annual Conference & Exposition, Minneapolis, MN. https://peer.asee.org/40487
Publisher's Copyright and Source
Copyright © 2022 American Society for Engineering Education.
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: © 2022 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.
Databases and Information Systems Commons, Educational Assessment, Evaluation, and Research Commons, Higher Education Commons