Scaffolding AI Research Projects Increases Self-efficacy of High School Students in Learning Neural Networks (Fundamental)

S. Shailja; Satish Kumar; Arthur Caetano; Ayush Pandey

Download Paper | Permalink

Conference: 2024 ASEE Annual Conference & Exposition
Location: Portland, Oregon
Publication Date: June 23, 2024
Start Date: June 23, 2024
End Date: July 12, 2024
Conference Session: Mr. Burns' Brainchild: AI in the Springfield STEM Classroom, Release the Hounds!
Tagged Division: Pre-College Engineering Education Division (PCEE)
Permanent URL: https://strategy.asee.org/47953

Request a correction

Paper Authors

biography

S. Shailja University of California, Santa Barbara orcid.org/0000-0002-5056-9989

visit author page

Shailja is an incoming post-doctoral fellow at Stanford University. She completed her Ph.D. in the Electrical and Computer Engineering (ECE) Department with interdisciplinary emphasis on College and University teaching at the University of California, Santa Barbara (UCSB) in 2024. She graduated with a bachelor's degree from the Electrical Engineering Department at the Indian Institute of Technology, Kharagpur in 2016. Shailja has been awarded the Fiona and Michael Goodchild best graduate student mentor award during her PhD. She has also been named an NSF iRedefine ECE Fellow for leadership potential among underrepresented graduate students across US/Canada. Shailja’s research vision is to develop AI methods for healthcare that “close-the-loop” between surgeons, research scientists, educators, and engineers.

visit author page

biography

Satish Kumar University of California, Santa Barbara

visit author page

Machine Learning Researcher-PhD student at UC Santa Barbara with 10+ years of research experience building advanced algorithms for large-scale solutions. It includes 6+ years in computer vision and machine learning algorithms and infrastructure at Vision Research Lab at UCSB. The current research is on multi-spectral image analytics, and I lead the project BisQue, an open source ML platform for data storage, AI/ML analysis, and visualization.

visit author page

biography

Arthur Caetano University of California, Santa Barbara

visit author page

Arthur Caetano is a Human-Computer Interaction Ph.D. student at the University of California, Santa Barbara, researching generative user interfaces in Extended Reality at the Human-AI Integration Lab under Prof. Misha Sra. With a Bachelor of Science in Computer Science from Universide Federal Fluminense (2017), he brings 5 years of experience in Product Management within the financial industry, focusing on internal technical solutions for data scientists and data platform regulators. Arthur also mentors high-school and undergraduate students in research and has 2 years of teaching assistant experience in Human-Computer Interaction and Computer Graphics.

visit author page

biography

Ayush Pandey University of California, Merced orcid.org/0000-0002-2524-2618

visit author page

Ayush Pandey is an Assistant Professor of Teaching in the Electrical Engineering department at the University of California, Merced. Before joining UC Merced, he completed his Ph.D. in Control and Dynamical Systems at the California Institute of Technology working with Dr. Richard M Murray. He is interested in increasing access to computational tools for inclusive data analysis education and research. His dissertation research develops computer-aided design tools for the engineering of biological systems to address sustainability and health challenges at scale. In 2022, he was appointed as an Adjunct Professor at the Harvey Mudd College where he explored the role of mathematical modeling and analysis tools in interdisciplinary computational biology education. Prior to that, in 2018, he obtained his master’s degree in Electrical Engineering from the California Institute of Technology. In 2017, he obtained dual bachelor’s and master’s degrees in the Electrical Engineering department from the Indian Institute of Technology (IIT), Kharagpur.

visit author page

Download Paper | Permalink

Abstract

With the rise of Artificial Intelligence (AI) in the mainstream and the impending need for an AI-trained workforce, we must devise strategies to lower the entry barrier to AI education. Advanced mathematical preparation and computational thinking skills are two major barriers in imparting a rigorous AI course at the high school level. Consequently, many existing AI-focused educational programs for high school students are basic primers and lack technical depth. In this paper, we assess two pedagogical instruments for increasing the self-efficacy of students in learning neural networks at the high school level. The first research question is whether high school students learn the basics of neural network design through scaffolded AI research projects. We also explore whether a dual advising structure with a research mentor and a communication teaching assistant enhances student’s self-efficacy in computing. For both of these questions, we define key variables to quantify student mastery and their computational thinking using qualitative student feedback and student reflection using GPT-3. We provide a reproducible blueprint for using large language models in this task to assess student learning in other contexts as well. We also correlate our results with a pre- and post-course Likert survey to find significant factors that affect student self-efficacy and belonging in AI.

With our course design and dual advising mentoring model, we find that students showed a significant improvement in their ability to articulate technical aspects within the AI domain and an increase in their confidence in speaking up in the AI field. Two out of the ten research projects applied AI techniques beyond classroom teachings, yielding original research contributions, and another six showcased students' capabilities in building neural networks from scratch. Our study has a strong selection bias since it focuses on top-performing students. However, the exploration of the two pedagogical instruments (scaffolding research projects and dual advising structure) aimed at high school students provides promising insights for future AI curricula design at the high school level.

Citation
Format

Shailja, S., & Kumar, S., & Caetano, A., & Pandey, A. (2024, June), Scaffolding AI Research Projects Increases Self-efficacy of High School Students in Learning Neural Networks (Fundamental) Paper presented at 2024 ASEE Annual Conference & Exposition, Portland, Oregon. https://strategy.asee.org/47953

TY - CPAPER
AB - With the rise of Artificial Intelligence (AI) in the mainstream and the impending need for an AI-trained workforce, we must devise strategies to lower the entry barrier to AI education. Advanced mathematical preparation and computational thinking skills are two major barriers in imparting a rigorous AI course at the high school level. Consequently, many existing AI-focused educational programs for high school students are basic primers and lack technical depth. In this paper, we assess two pedagogical instruments for increasing the self-efficacy of students in learning neural networks at the high school level. The first research question is whether high school students learn the basics of neural network design through scaffolded AI research projects. We also explore whether a dual advising structure with a research mentor and a communication teaching assistant enhances student’s self-efficacy in computing. For both of these questions, we define key variables to quantify student mastery and their computational thinking using qualitative student feedback and student reflection using GPT-3. We provide a reproducible blueprint for using large language models in this task to assess student learning in other contexts as well. We also correlate our results with a pre- and post-course Likert survey to find significant factors that affect student self-efficacy and belonging in AI.

With our course design and dual advising mentoring model, we find that students showed a significant improvement in their ability to articulate technical aspects within the AI domain and an increase in their confidence in speaking up in the AI field. Two out of the ten research projects applied AI techniques beyond classroom teachings, yielding original research contributions, and another six showcased students&#39; capabilities in building neural networks from scratch. Our study has a strong selection bias since it focuses on top-performing students. However, the exploration of the two pedagogical instruments (scaffolding research projects and dual advising structure) aimed at high school students provides promising insights for future AI curricula design at the high school level.
AU - S. Shailja
AU - Satish Kumar
AU - Arthur Caetano
AU - Ayush Pandey
CY - Portland, Oregon
DA - 2024/06/23
PB - ASEE Conferences
TI - Scaffolding AI Research Projects Increases Self-efficacy of High School Students in Learning Neural Networks (Fundamental)
UR - https://strategy.asee.org/47953
ER -