Awarded Pilot Projects for 2020-2021

Using Conversational Agents To Support Older Adult Learning For Health Morrow-etal-Figure 2 final

The research team’s goal is to develop and evaluate a Conversational Agent (CA) system to support self-care. Interactive CAs in particular have the potential to present health information tailored to older adults when and where needed, thus expanding opportunity for patient education in a health care system where providers do not have the time to effectively educate their patients. More broadly, CA systems may support independent living and work at home in times of public health crises such as the current COVID-19 pandemic. The project aims to conduct a follow-up study to further investigate the potential of CA-based teachback to support older adult learning about self-care and also refine the interactive capabilities of the CA system.

The team combines expertise in CA speech processing (Mark Hasegawa-Johnson, ECE Dept), Natural Language Processing (Suma Bhat, ECE Dept), behavioral-social science (Dan Morrow, Educational Psychology Dept), and medicine (James Graumlich, UIC Medical School and OSF Health System). They have worked together for several years to develop and begin to evaluate the current CA prototype.

Supporting Self-Regulated Learning In Online Education Via Automatically Personalized Interventions

This research team is proposing a generalizable, interdisciplinary approach to provide individualized Self-Regulated Learning (SRL) training in online courses as a means of addressing the growing need to improve online learning and the issues of equity that go with it. In particular, they propose to integrate research on SRL with machine learning to intelligently select and display interventions designed to promote the SRL skills most needed by a particular student at the right time. Their method aims to provide personalized SRL interventions in large-scale online courses, utilizing machine learning to predict which students need intervention (via student outcome prediction) and what that intervention should be (via Shapley analysis of the model’s prediction for that student).

The project team includes experts in machine learning, online education, and in-depth human subjects research, all of which are essential to this interdisciplinary project. Nigel Bosch has experience in research on self-regulated learning, machine learning, and equity for students from traditionally-underrepresented groups in digital learning environments. Bosch will serve as principal investigator, focusing especially on SRL feature design and web development aspects of the project. Suma Bhat has expertise in human–computer interaction, machine learning, and analysis of online STEM learning environments. Bhat will serve as co-principal investigator, focusing especially on machine learning and natural language processing. Paul Hur is a PhD student who has experience conducting in-depth interviews with participants in human–computer studies. Hur’s expertise will be particularly valuable during data collection, when we will be interviewing participants and revising the pilot study interface.

Propelling Teacher Professional Development Through FAAST Feedback On Student Epistemic Views

This project proposes to implement a feedback system for formative, automated assessment of student thinking (FAAST) and to use this feedback to fuel teacher professional development. The FAAST feedback system uses a blend of machine-learning techniques and human-driven inductive coding to provide immediate feedback to students and teachers on classroom-level patterns in thinking. Currently, the FAAST system has been designed to categorize epistemic thinking about how general or specific scientific explanations should be. This project will use the FAAST system to support teacher reflection on epistemic learning goals in science.

This team is led by Stina Krist, an assistant professor in Curriculum & Instruction at the University of Illinois at Urbana-Champaign. Her research focuses on students’ epistemic views and practices in K-12 science learning settings, as well as how teachers learn to create and sustain environments that foster students’ epistemic agency. Eric Kuo is one of the Co-PIs and is an assistant professor in Physics at the University of Illinois at Urbana-Champaign. His research focuses on students’ epistemic views in physics, and how they inform the learning and application of problem-solving strategies in undergraduate courses. The other Co-PI is Joshua Rosenberg who is an assistant professor of STEM Education at the University of Tennessee, Knoxville. His research focuses on student engagement across a range of STEM learning environments and the use of technology in education, particularly for teacher professional development. He led the initial development of the FAAST feedback system for epistemic views of generality.

Interactive Overlays In Multimedia For Science Learning: Attention And Comprehension

This research will explore the relative effects of adjuncts and textual labels as interactive overlays in multimedia for science learning. To what extent do virtual adjuncts and configurations enhance visual cognition, resource availability, and meaning comprehension? To what extent do students utilize these features while consuming multimedia for science learning, and how do their effects unfold dynamically over time? The answers to these questions have the potential to broadly impact both the understanding and design of educational technology across a wide range of learning domains (e.g. biology, ecology, physics) and, to a lesser extent, individual differences (e.g., age, expertise, socioeconomic status) among users of such technology. This project will help to realize the potential of the taxonomy for improving scientific visualization, providing experimental evidence of how cognitive image functions impact the interpretational processes that lead to learning.

Kevin Wise, Professor of Advertising is the PI of this team. The Co-PI is Matthew Peterson, Assistant Professor of Graphic Design at North Carolina State University. Xiaoyu Xu is a PhD student (Institute of Communications Research) who will oversee study implementation, data collection, and analysis. Another graduate student will join the team to produce illustrations in a format that enables development of the interactive multimedia stimuli.

 

TIER-ED Graduate Fellows for 2020-2021

Gerry Derksen is originally from Canada where he went to the University of Manitoba’s architecture school and later finished his master’s degree at the University of Alberta. Under Jorge Frascara he studied visual communication design, which integrated user-centered design philosophy with traditional marketing communication strategies. Much of Gerry's published research deals with human behavior, cognition and perceptions of interactive media, visualization, and designed experiences. Currently, he is working at Winthrop University in Digital Information Design active in the practice and research of user experience design, usability and visual media. The subject of his Ph.D. dissertation, “Educational Smart Toys for Autistic Children”, examines the use of machine learning to identify patterns in learning to aid students with disabilities. Educational game developers understand the importance of relating content to game action to reinforce learning. With the advent of artificial intelligence, the use of neural networks has the potential to predict human behavior. Capturing these physical interactions as well as more traditional assessments of progress, maps the learning process teaching the intelligent toy to provide support if it predicts struggle. In this study, artificial intelligent toys attempt to record the patterns of learning identifying moments best suited for autistic students to learn, mimic, transition, and apply new information.

Taehyun Kim is a second-year doctoral student in the Digital Environments for Learning, Teaching & Agency (DELTA) in the department of Curriculum and Instruction. He is interested in designing emerging media platforms such as simulations, virtual and mixed reality, or video games with the framework of embodied cognition and design-based research. He seeks to understand how these technologies can be used to help learning and teaching situations, especially in STEM domain. His research focused on how the body-based interactions with the designed media can facilitate complex understanding of learning contents, and how the media can be adequately designed to include these types of interactions. Along with these research interests, this TIER-ED funded project aims to investigate the effects of differing levels of embodied interactions have on students’ science learning in an immersive VR learning environment. This project explores how learners use their bodies and gestures to represent the scientific phenomenon they are trying to understand, and how embodied VR simulations can support students’ learning about the cell division process.

taehyun

Graduate Research Award 2020-2021

James

James Lee is a doctoral student in the Department of Special Education. James has worked as a clinician in the field of applied behavior analysis (ABA) in California and South Korea. His experiences include directing ABA clinic operations and conducting research in low-resource settings. His research interests focus on providing training and coaching with parents of young children with autism in low-resource settings and collaborating with international stakeholders in response to the scarcity of resources in many parts of the world. James’ dissertation topic is on training and coaching parents of children with autism in Mongolia with NGO staff as the intermediary using telehealth technology in a single-case research. His study will focus on building capacity within a community of parents of children with autism in a low-resource setting in a cascading intervention model.

Awarded projects relevant to TIER-ED focus and goals:

The Role of Online Museum Experiences in Supporting At-Home Science Learning in the Era of COVID-19: 

  • Catherine Dornfeld Tissenbaum, Curriculum & Instruction, PI; Stina Krist, Curriculum & Instruction, Co-PI; McKenna Lane, Curriculum & Instruction, Graduate Student.

COVID-19 Risk Mitigation: Interactive Automatic Counselor

  • Dan Morrow, Educational Psychology, PI; H. Chad Lane, Educational Psychology, Co-PI; Suma Bhat, Electrical and Computer Engineering, Co-PI; Mark Hasegawa-Johnson, Electrical and Computer Engineering, Co-PI.

City Settlers: Transforming a Collaborative STEM Education Game for COVID-19 Online Use

  • Mike Tissenbaum, Curriculum & Instruction, PI; Vikesh Kumar, University of Wisconsin, External Consultant; Litong Zheng, Curriculum & Instruction, Graduate Student; Ruby Wang, College of Engineering, Undergraduate; Taehyun Kim, Curriculum & Instruction, Graduate Student; Zhanchen Huang, School of Information Sciences, Undergraduate.

 

TIER-ED Researchers Re-thinking Digital Exhibits in Museum Spaces

Several projects are currently active where TIER-ED faculty and students are re-designing the ways that museums engage their visitors with digital content.

In one project, TIER-ED Director Robb Lindgren is taking his work on embodied learning simulations onto the museum floor such that young science center visitors can use the full movement of their bodies to interact with a game about energy. The game, called energIze uses the Microsoft Kinect to empower children to act out different forms of energy, energy transfer, and to reason about conservation of energy. Dr. Lindgren and his TIER-ED colleague Dr. Mike Tissenbaum are currently working with the Lawrence Hall of Science in Berkeley, CA, where the energIze exhibit is currently active on the museum floor, to develop a system of connected embodied learning exhibits that build off of each other and facilitate learning transfer.

robb_proj
dsc_0477_2

In another project Dr. Emma Mercier is examining parent-child collaboration in an informal learning context. Along with graduate student LuEttaMae Lawrence, she is implementing their Food for Thought application onto a large screen surface at a local museum. Food for Thought focuses on using a familiar topic (food) to help students understand more about climate change, and the way small choices can have a global impact. The software was initially designed for use on large multi-touch table, and two iterations were tested with middle school students through the CoLearn Lab

Finally, Dr. H Chad Lane is on the leadership team for a major international collaboration with multiple museums in the U.S. and the U.K. understanding the role of gesture and other kinds of body movement for pre-school children’s science learning in informal environments. The consortium, which also includes TIER-ED Director Lindgren, is called Move2Learn, and it brings together researchers and museum practitioners to study how experiences such as the Rivers of Grass exhibit at the Frost Museum of Science in Miami, FL encourage students to use their bodies when engaging with science ideas. Other digital exhibit research on the Move2Learn project is happening at The Children’s Museum’s Indianapolis where young visitors use their whole bodies to learn critical ideas in crop sciences.

NSF-funded Projects Create New Digital Environments for Learning

TIER-ED community members have been very successful in receiving funding from external agencies such as the National Science Foundation.

One such project is the GRASP simulation platform that was co-designed by UofI College of Education researchers. The goal of GRASP (GestuRe Augmented Simulations for supporting exPlanations) is to understand the role gestures play in reasoning about critical concepts in science, and to create simulations that encourage the use of gestures for creating scientific explanations. GRASP is a NSF-funded collaboration between UofI and Concord Consortium, an educational research and development organization in Massachusetts. Robb Lindgren (PI) and David Brown (Co-PI) are working with graduate students to interview local middle school students to understand how the gesture-augmented simulations affect their science explanations and understanding.  

 

grasp-hands-3
scholar2

In a second project, Dr. Bill Cope and Dr. Mary Kalantzis explore ‘complex epistemic performance’ in online learning environments as part of an NSF-funded project to further develop Common Ground Scholar (CGScholar) environment. They have launched Scholar’s new Analytics area, created by their research team at the University of Illinois. Analytics is a learning visualization tool, wherein any unit of work a student can see their progress towards mastery, and a teacher can see the comparative progress towards mastery of all members of the class, identifying which students may require more time or special attention to achieve mastery. The CGScholar Analytics tool collects information about learning processes in three domains: knowledge, focus, and collaboration. Developing this tool is an interdisciplinary effort involving faculty and graduate students from multiple units across campus.