NSF SaTC Project: Pilot Test on Saturday, June 25, 2022
June 25, 2022
We pilot-tested the Private AI curriculum at the conference room, 25 Park Place, on Saturday, June 25, 2022. 29 undergraduate and graduate CS students participated in the pilot test. Our two project GRAs, Prajwal Panzade and Lia Haddadian, developed the test module, Privacy-Preserving Machine Learning (PPML), using Differential Privacy with the focus, Differential Privacy & TensorFlow. Akshita Maradapu Vera Venkata sai, a 4th-year Ph.D. student in the Department of Computer Science at Georgia State University, was recruited to lecture the module.
The CS curriculum “Private AI” consists of 10 modules that teach students to use specific techniques to address various privacy challenges in AI systems. We designed each module based on the principles of Problem-Centered Instruction (PCI) that feature a series of learning activities: real-world problem scenarios, instructor-led instruction, individual assignments supported by worked examples, crowd-based hands-on lab activities in pairs, and debriefing lessons learned. The data collection includes pre-and post-test, a survey, video recordings of hands-on lab activities, audio recordings of the debriefing session, instructor interview, and student interviews.
Min Kyu Kim and Nam Ju Kim presented at the 16th International Conference of the Learning Sciences
June 9, 2022
Co-directors, Drs. Min Kyu Kim and Nam Ju Kim presented a short paper at the 16th International Conference of the Learning Sciences on June 9th, 2022. Due to the pandemic, the ICLS meeting was held virtually. They had developed the initial work into two journal publications since the proposal submission to ICSL. (This post includes two published articles associated).
Kim, M., & Kim, N. (2022). AI-supported scaffolding for writing academic arguments. In C. Chinn, E. Tan, C. Chan, & Y. Kali (Eds.), Proceedings of the 16th International Conference of the Learning Sciences-ICLS2022 (pp. 1129-1132). Hiroshima, Japan: International Society of the Learning Sciences.
Despite the importance of writing academic arguments, instructors often lack enough time and knowledge to provide prompt support adaptive to individual students' discipline-specific arguments. AI techniques can enable automated and adaptive educational scaffolding. In this study, we created a test version of the Artificial Intelligent-Supported Scaffolding (AISS) system that provides scaffolds in the form of alternative writing examples that human experts would write. Our mix-methods data gathered from 14 students enrolled in two sections of the same graduate-level online course revealed that the students leveraged AI-generated scaffolds to build a more substantial claim with elaborated ideas in a cohesive text structure. The current study's findings demonstrated the potential of AI to provide personalized scaffolding for writing academic arguments.
Published Counterparts:
Kim, N., & Kim, M. (2022). Teacher's perceptions of using an artificial intelligence-based educational tool. Frontiers in Education, 7(755914), 1-13. https://doi.org/10.3389/feduc.2022.755914
Kim, M., Kim, N. & +Heidari, A. (2022). Learner experience in Artificial Intelligence scaffolded argumentation. Assessment and Evaluation in Higher Education. Advance online publication. DOI: 10.1080/02602938.2022.2042792
Dr. Min Kyu Kim attended the AI-ALOE External Advisory Board Meeting
April 26, 2022
Dr. Min Kyu Kim attended the AI-ALOE External Advisory Board Meeting at Georgia Tech on April 26, 2022. It was the first in-person meeting of the AI-ALOE Institute as a whole team! At the meeting, Dr. Kim also presented a recent study conducted with our co-director, Dr. Nam Ju Kim.
A Design Study of the Artificial Intelligence-Augmented Motivation Indicator (AIMI) System
In online learning, motivation encourages learners to initiate certain actions, use appropriate strategies to achieve their goals, and sustain actions even in difficult environments. Therefore, detecting students’ motivation accurately and on time is crucial as educators can easily identify what learning contents, activities, and instructional methods significantly undermine learners’ motivation in the online learning environment. To measure students’ real-time motivation level in online learning, we developed Artificial Intelligence-Augmented Motivation Indicator (AIMI) system. This study aims to validate the accuracy of AIMI by comparing motivation levels measured by both this tool and traditional paper-based surveys. As a result, the motivation values generated by the AIMI system demonstrated a high level of accuracy, with an error rate of only about 10%.
Presentations at 2022 AERA Annual Meeting
April 25, 2022
Our co-directors, Drs. Min Kyu Kim and Nam Ju Kim, presented two posters at 2022 Annual Meeting for American Educational Research Association (AERA) in April 21-26, 2022, San Diego.
Revisiting a Three-Stage Learning Progression Model Through a Technology-Based Formative Assessment System (Poster 5)
Fri, April 22, 4:15 to 5:45pm PDT (7:15 to 8:45pm EDT), SIG Virtual Rooms, SIG-Instructional Technology Virtual Poster Session Room
Link to i-Presentation
In a previous study, we proposed a three-stage learning progression model that drew on theories of mental models and expertise development to test stages through which students develop an expert-like understanding of a problem situation. Even though the initial work showed promising results, the student data were manually processed and thus required further validation with computerized data analytics. This study revalidated the learning progression model, using data generated by a technology-based formative feedback system. We used textual explanations about a complex problem scenario written by 136 students and 6 experts. Descriptive statistics and Exploratory Factor Analysis (EFA) results demonstrated the existence of latent knowledge attributes. We are now in further investigation, and full results will be presented at the convention.
A Design Study of the Artificial Intelligence-Augmented Motivation Indicator System (Poster 1)
Mon, April 25, 2:30 to 4:00pm PDT (5:30 to 7:00pm EDT), San Diego Convention Center, Floor: Upper Level, Sails Pavillion
In online learning, motivation encourages learners to initiate certain actions, use appropriate strategies to achieve their goals, and sustain actions even in difficult environments. Therefore, detecting students’ motivation accurately and on time is crucial as educators can easily identify what learning contents, activities, and instructional methods significantly undermine learners’ motivation in the online learning environment. To measure students’ real-time motivation level in online learning, we developed Artificial Intelligence-Augmented Motivation Indicator (AIMI) system. This study aims to validate the accuracy of AIMI by comparing motivation levels measured by both this tool and traditional paper-based survey. As a result, the motivation values generated by the AIMI system demonstrated a high level of accuracy, with an error rate of only about 10%.
Crystal Bundrage presented at the 32nd International SITE Conference
April 13, 2022
Our graduate research associate, Crystal Bundrage, presented at this year’s Society for Information Technology & Teacher Education Conference. Their presentation, “Design and Development of an Online Professional Development Course on Culturally Responsive Pedagogy Using the ADDIE Model”, described the process of creating an asynchronous institutional, professional development course centered on Culturally Responsive Pedagogy (CRP). CRP uses cultural knowledge, and the lived experiences of ethnically diverse students as a conduit for teaching more effectively. As a more diverse population grows, there is a need for educators to improve their intercultural fluency and to incorporate it into their pedagogy.
Bundrage, C. & Mapson, K. (2022). Design and Development of an Online Professional Development Course on Culturally Responsive Pedagogy Using the ADDIE Model. In E. Langran (Ed.), Proceedings of Society for Information Technology & Teacher Education International Conference (pp. 248-256). San Diego, CA, United States: Association for the Advancement of Computing in Education (AACE). Retrieved April 29, 2022 from https://www.learntechlib.org/primary/p/220742/.
AI-ALOE Introduction Video
April 11th, 2022
Dr. Chris Dede, one of the co-PIs of the AI-ALOE institute, created a fantastic video introduction to AI-ALOE. This video was submitted to the 2022 STEM for ALL Video Showcase. This video gives you an overview of the National AI Research Institute for Adult Learning and Online Education (AI-ALOE). Funded by the National Science Foundation and headquartered at Georgia Tech, AI-ALOE aims to leverage AI technologies for transforming adult education in effectiveness, efficiency, scale, and personalization, thereby making it more available, achievable, and equitable.
Runner-Up Best Research Paper Award
March 23, 2022
“A Design Study of the Artificial Intelligence Motivation Indicator System” coauthored by co-directors Dr. Nam Ju Kim and Dr. Min Kyu Kim, is the Runner-up for the Online Teaching and Learning SIG 2022 Best Research Paper Award.
Motivation is a critical component of effective learning and directly affects academic achievement. In this study, we have developed the Artificial Intelligence-Augmented Motivation Indicator (AIMI) system that is tuned to detect learners’ real-time motivation levels to help students sustain their motivation for learning activities. The AIMI system’s unique approach leverages artificial intelligence techniques to analyze learners’ facial expressions and determine their emotions and associated motivation levels in real-time. A pilot version of AIMI demonstrated high accuracy of motivation detection, which supports the potential use of Artificial intelligence in the field of psychology and education.
Publication on teachers’ perceptions of AISS
March 29, 2022
Kim, N., & Kim, M. (2022). Teacher's perceptions of using an artificial intelligence-based educational tool. Frontiers in Education, 7(755914), 1-13. https://doi.org/10.3389/feduc.2022.755914
Abstract
Efforts have constantly been made to incorporate AI into teaching and learning; however, the successful implementation of new instructional technologies is closely related to the attitudes of the teachers who lead the lesson. Teachers’ perceptions of AI utilization have only been investigated by only few scholars due to an overall lack of experience of teachers regarding how AI can be utilized in the classroom as well as no specific idea of what AI-adopted tools would be like. This study investigated how teachers perceived an AI-enhanced scaffolding system developed to support students’ scientific writing for STEM education. Results revealed that most STEM teachers positively experienced AI as a source for superior scaffolding. On the other hand, they also raised the possibility of several issues caused by using AI such as the change in the role played by the teachers in the classroom and the transparency of the decisions made by the AI system. These results can be used as a foundation for which to create guidelines for the future integration of AI with STEM education in schools, since it reports teachers’ experiences utilizing the system and various considerations regarding its implementation.
Publication on AISS
February 14, 2022
Kim, M., Kim, N. & Heidari, A. (2022). Learner experience in Artificial Intelligence scaffolded argumentation. Assessment and Evaluation in Higher Education. Advance online publication. http://doi.org/10.1080/02602938.2022.2042792
Abstract
Writing academic arguments is a complex and demanding task, even for proficient tertiary students. At the same time, providing prompt support for individual students working on discipline-specific arguments is often challenging for instructors. Artificial Intelligence (AI) techniques enable automated and adaptive educational scaffolding. In this study, we leveraged Natural Language Processing (NLP) AI techniques in an AI-Supported Scaffolding (AISS) system to evaluate written arguments and present alternative writing examples that human experts might write. To evaluate a pilot version of AISS, we gathered mixed-method data from 14 students enrolled in two sections of the same graduate-level online course (6 students in Cohort 1 and 8 students in Cohort 2). We used the Tool for the Automatic Analysis of Cohesion (TAACO) to track revisions in written arguments. TAACO indices demonstrated that the students used AI-generated scaffolding to build stronger claims with more elaborate and cohesive ideas. In written reflections, students revealed their perceived value of AISS, and visual inspection of their written arguments indicates that students used AISS feedback to improve their arguments. The findings demonstrate the potential of AI to provide personalized scaffolding for academic argument composition.
Keywords: Artificial intelligence, scaffolding, expert model, academic argument, personalized learning
Yoojin Bae joins as a graduate research associate
January 1, 2022
Welcome Yoojin to the AI2 research laboratory!
Yoojin is a Graduate Research Associate and a Ph.D. student at the Department of Learning Sciences in the College of Education and Human Development, Georgia State University (GSU). She earned her master’s degree in Education from Seoul National University, South Korea. Her research interests lie in the intersection of Education and Computer Science. Her work’s focus is on examining ways to leverage artificial intelligence in order to offer adaptive learning environment. Yoojin is exploring natural language processing to analyze text data in educational circumstances.