2025 ASEE Annual Conference & Exposition

Natural Language Processing Models to Detect Affective Fluctuations of Engineering Faculty and Students Responding to a Hidden Curriculum Survey

Presented at ERM: Faculty Influences on Student Support

This empirical research full paper considers a validated mixed-methods vignette UPHEME (Uncovering Previously Hidden Engineering Messages for Empowerment) survey (2018-2019). For this study, 961 participants who belonged to a diverse population of engineering students and faculty were surveyed in Spring 2019. The survey was divided into four categories including hidden curriculum awareness, emotions, self-efficacy, and self-advocacy. The participants answered a few questions on hidden curriculum (HC) before they watched the video, then responded to some more questions on HC, self-identified their own emotion from 14 emotion categories and classified it as a positive or negative emotion on a 5-point Likert scale as they transitioned to later sections of the survey. This study uses Natural Language Processing (NLP) to analyze the responses of participants for one open-ended question in the ‘emotions’ category using four different pre-trained models from HuggingFace platform, an open-source platform for machine learning and data science, for detection of at least six emotions (sadness, joy, anger, surprise, fear, love and/or neutral). This study provides a contrast of the emotions experienced by engineering participants for all four models, explores the identified emotions across the demographics (primary discipline, self-identified age range, self-identified gender identity, self-identified race/ethnicity) of participants and compares the models with each other. The findings reflect a range of emotions as identified by four models and the need for an intersectional approach in developing inclusive strategies with a cultural and emotional awareness to empower individuals in navigating academic and professional settings. There are several implications of the study including how the participants’ awareness of hidden curriculum affects their emotions which in turn affects their self-efficacy and self-advocacy, and their demographic correlations. It gives us insights to utilize NLP techniques for qualitative data within a mixed-methods survey to extract meaningful information for educational research in engineering education.

Authors

Mr. Gadhaun Aslam University of Florida [biography]

Gadhaun Aslam is a PhD Candidate & Graduate Assistant in the TWISTER Lab within the Department of Engineering Education at University of Florida (UF). His research interests include extracting data from institutional websites to understand the trajectory of engineering education, exploring student learning experiences using multi-modal tools (e.g., eye tracking and physiological electrodermal sensors) and integrating data science and machine learning in understanding the links between cognition, motivation, and performance in STEM classrooms and connected activities, and improving faculty mentoring experiences. He is actively involved in graduate student organizations. He is serving as President of Global Connections (GC) and Engineering Education Graduate Student Council (EEd GSC) at UF, and as Student Division Information Chair at the American Society of Engineering Education (ASEE).
Yuxuan Wang University of Florida [biography]

This is Yuxuan from University of Florida. I am a graduated master student of UF Computer & Information Science & Engineering department. I am currently conducting an independent study research work at Dr. Idalis Villanueva Alarcon's leading Twister Lab of UF Engineering Education Department.
Idalis Villanueva Alarcón http://orcid.org/https://0000-0002-8767-2576 University of Florida [biography]

Dr. Idalis Villanueva Alarcón is Chair and tenured Associate Professor in the Department of Engineering Education in the Herbert Wertheim College of Engineering. A PECASE awardee, she has led multiple pioneering efforts in engineering education including multimodal methods in engineering education using sensor technologies and biophysiological tools, hidden curriculum, mentoring, active learning, professional identity, among others. She is a renowned national and international leader in engineering education earning her multiple accolades and honors through professional organizations such as the National Academy of Engineering, IEEE, and ASEE. She integrates her multiple experiences as a Chemical Engineering, Biological Engineer, Analytical Cell Biologist, and Engineering Education Researcher to tackle complex engineering education problems across the learner life span.
Dr. Edwin Marte http://orcid.org/0000-0002-8700-3882 University of Florida [biography]

Dr. Edwin Marte Zorrilla is an Instructional Assistant Professor in the Department of Engineering Education at the University of Florida. His research interests include computational and technology-driven engineering education, learning analytics with multimodal tools, and advancements in learning technologies.

Note

The full paper will be available to logged in and registered conference attendees once the conference starts on June 22, 2025, and to all visitors after the conference ends on June 25, 2025

« View session

For those interested in:

Academia-Industry Connections
Advocacy and Policy
engineering
Faculty
gender
Graduate
race/ethnicity
undergraduate