Key Moments
Research investigating the science and engineering of human learning
MIT OpenCourseWareKey Moments
MIT researchers present diverse projects on educational science and engineering, focusing on learning, engagement, and skill development.
Key Insights
Predictive models can identify learners at risk of dropping out of MOOCs, with content-focused interventions proving more effective than motivational ones.
Integrating computational tools like coding and visualizations into introductory physics courses shows promise but requires careful experimental design, especially with ongoing enrollments.
Real-time biofeedback, especially through wearable devices like "Attentiview," can monitor and potentially improve attention and learning performance, though hardware development faced pandemic challenges.
Adaptive tools that dynamically adjust difficulty levels, such as in motor skill training, can lead to higher learning gains than static approaches, as learners often misjudge their own skill levels.
Leveraging digital game assets to create fabrication files for maker skills offers a novel way to teach fabrication and maker skills to children, with potential applications in STEM education.
Novelty insertion into learning environments may enhance plasticity and improve learning outcomes, though the distinction between novel and familiar stimuli can be subtle and requires careful consideration.
AI-powered social agents can support parent-child interactions during story time, enhancing conversational turn-taking and language development, with potential for personalized interventions.
Virtual reality (VR) environments, like the "CellVerse" game, can significantly improve understanding of complex biological concepts and enhance spatial awareness compared to 2D interfaces.
REDUCING DROPOUT AND ENHANCING MOOC LEARNING
Research focused on Massive Open Online Courses (MOOCs) aimed to tackle the significant challenge of learner dropout. A predictive model using random forest and logistic regression demonstrated high accuracy in identifying students likely to drop out, primarily based on grades and time spent on the course. While motivational interventions like emails had no effect, interventions focused on improving course content, such as gradually increasing difficulty and providing better support materials, showed a positive impact in reducing dropout rates, as evidenced by successful publications and conference presentations.
COMPUTATIONAL TOOLS IN INTRODUCTORY PHYSICS
This project investigated the impact of integrating computational elements, specifically coding and interactive visualizations, into introductory online physics courses. While the study did not yield statistically significant results, it provided valuable insights into experimental design challenges in online learning environments, such as managing on-the-fly cohorting for randomized groups. The data suggested potential, albeit not statistically significant, benefits of programming and visualizations for physics and spatial comprehension, highlighting the need for further exploration and refined experimental methodologies.
BIOFEEDBACK FOR SUSTAINED ATTENTION
The 'Attentiview' system, a pair of glasses equipped with EEG and EOG sensors, was developed to measure real-time brain activity and eye movements for biofeedback. The research aimed to evaluate the effectiveness of this closed-loop biofeedback system in enhancing sustained attention during actual homework tasks, particularly for children and young adults. Despite pandemic-related hardware supply chain issues, pilot studies and pre-study results indicated that personalized calibration and real-time feedback could potentially improve learning outcomes and engagement, with a focus on objective performance metrics and subjective feedback.
ADAPTIVE LEARNING FOR MOTOR SKILLS
This research explored the concept of adaptive tools for learning motor skills, exemplified by an adaptive basketball hoop that adjusted height and width based on player performance. User studies demonstrated that learners often struggle to accurately self-assess their skill levels, leading to suboptimal training difficulty. Auto-adaptive modes consistently showed higher learning gains compared to manual or non-adaptive conditions, fostering trust and reducing learner anxiety. The project also included workshops that encouraged students to build adaptive tools, leading to the development of user interfaces and visualization tools for designing and understanding adaptive learning systems.
FABRICATION SKILLS THROUGH GAMING
This project focused on developing a system called 'Fabo' to enable the fabrication of in-game objects, thereby teaching children fabrication and maker skills. By capturing digital assets from video games, the system generates fabrication files for technologies like laser cutting, allowing players to create physical versions of their in-game items. User studies indicated that this approach is highly engaging and has the potential to excite children about STEM fields. Ongoing work involves interviewing educators to identify pain points and co-designing games to integrate fabrication learning seamlessly into gameplay.
NOVELTY INSERTION FOR ENHANCED PLASTICITY
The research hypothesized that immersing students in novelty could upregulate neuroplasticity and enhance learning outcomes. An experiment in a MOOC involved interspersing lecture material with either novel video clips (e.g., animals in the wild) or familiar ones (e.g., cleaning tips). While physiological responses to novel and familiar videos were unexpectedly similar among MIT students, subjects performed marginally better on material presented shortly after novel insertions, suggesting a potential, albeit subtle, effect of novelty on immediate learning.
DIALOGIC READING WITH SOCIAL AI AGENTS
This project aims to improve parent-child conversational turn-taking during story reading using social AI agents. Leveraging collected parent-child conversation data, the research developed affect prediction algorithms and explored different roles for an AI agent—bystander, question-asker, or suggestion-maker. Pilot studies, including teleoperated sessions with robots, provided insights into how AI can support parents, especially those for whom English is a second language, to foster more beneficial interactions and language development in children. Future work includes developing personalized agent policies tailored to specific parent-child dynamics.
VIRTUAL REALITY FOR CELL BIOLOGY EDUCATION
'CellVerse,' a collaborative VR and tablet-based game, immerses high school students in a cell with cystic fibrosis to learn cell biology concepts. Studies, including a randomized controlled trial comparing VR to a 2D version of the same game, showed that VR significantly improved players' understanding of cell environments and processes, enhanced spatial awareness, and led to higher content gains. The game's design emphasizes collaboration and interdependence between the VR 'explorer' and the tablet 'navigator,' making abstract biological concepts more concrete and engaging.
Mentioned in This Episode
●Products
●Software & Apps
●Organizations
●Books
●Concepts
●People Referenced
Common Questions
The Mightily research grants aim to enhance opportunities for researchers across MIT to improve research related to education and educational outcomes.
Topics
Mentioned in this video
Massachusetts Institute of Technology, where the Mightily research grants program supports researchers across various departments.
A research grant program supporting innovative research in education and learning outcomes across MIT.
Computer Science and Artificial Intelligence Laboratory at MIT, where Dashida is a PhD student researching skill learning.
Collaborator on the MOOC dropout reduction project, with ongoing availability for further questions.
Collaborator on the MOOC dropout reduction project.
Faculty director and professor in the department of brain and cognitive sciences, involved in organizing and presenting research supported by Mightily grants.
Professor at MIT CSAIL, supervising Dashida's PhD research on learning skills.
Collaborator on the MOOC dropout reduction project.
Mentioned in the closing remarks regarding future funding for research.
A machine learning technique used in the predictive model to identify learners likely to drop out of MOOCs.
A statistical method used in the predictive model for MOOC dropout prediction.
A game mentioned as an example where players could fabricate in-game items like swords.
A game mentioned where a uniquely designed skateboard was fabricated in user studies.
A game mentioned where fabricated objects, like an axe, were created in user studies.
A journal where the MOOC dropout reduction research is currently under second-round review for publication.
A MOOC by John Gabrieli used in the experiment on enhancing learning through novelty insertion.
A game mentioned where fabricated custom designs, like a dress, were created in user studies.
More from MIT Open Learning
View all 182 summaries
2 minWhy are nuclear power plants so expensive in the U.S.?
2 minThe science behind fake snow
1 minPaula Hammond: From curiosity to historic leadership at MIT
2 minGhost Trees Explained: What they reveal about climate change
Found this useful? Build your knowledge library
Get AI-powered summaries of any YouTube video, podcast, or article in seconds. Save them to your personal pods and access them anytime.
Try Summify free