Tuesday, March 4
Registration will be in the Regency Foyer, Ballroom Level LL1, of the Hyatt Regency Atlanta beginning at 7:30am. Breakfast will be served starting at 8am in the Regency V room. Paper sessions will take place in the Regency VII room. The poster/demo reception will be held in Regency V and the Terrace Foyer.
- Welcome from the PC Chairs:
Armando Fox, Marti Hearst, Michelene Chi
Talk Slides (pdf)
- Keynote Talk: New Wine in No Bottles: Immersive, Personalized, Ubiquitous Learning
Chris Dede (Harvard University)
Talk Slides (pdf)
- Session: Student Skills and Behavior
(Chair: Ben Bederson)
- Lunch (on your own)
- Session: Course Materials
(Chair: Doug Fisher)
- Session: The Role of the Instructor
(Chair: Karen Swan)
- Poster/Demo Introductions
- Poster/Demo Reception
Wednesday, March 5
Registration will be in the Regency Foyer, Ballroom Level LL1, of the Hyatt Regency Atlanta beginning at 7:30am. Breakfast will be served starting at 8am in the Regency VI room. Paper Sessions will take place in the Regency VII Room. The two tutorials will be held in parallel in Regency VI and VII.
- Panel: Online Learning Platforms and Data Science
(Moderator: Mehran Sahami).
Peter Norvig (Google), Andreas Paepcke (Stanford)
Talk slides (pdf), Jace Kohlmeier (Khan Academy) Talk slides (pdf), and Amin Saberi (NovoEd)
- Session: Assessment
(Chair: John Mitchell)
- Lunch (on your own)
- Session: Forums and Chat Rooms
(Chair: Kurt VanLehn)
- NSF Programs Relevant to Learning@Scale:
Janet Kolodner (NSF)
Talk slides (pdf)
- MOOCS in the Developing World; Lessons from India:
Ed Cutrell and Bill Thies (Microsoft Research India)
- Tutorial 1 (parallel)
- Tutorial 2 (parallel)
- Conference Close
Speaker: Chris Dede, Harvard University
Title: New Wine in No Bottles: Immersive, Personalized, Ubiquitous Learning
The invention of the movie camera was initially seen as a way to reach “massive” by filming plays; we are in an equivalent stage with our early MOOCs. Thinking outside the box of “teaching” is essential to realizing learning at scale. Virtual worlds and augmented realities can complement digitized classroom instruction through simulated apprenticeships, embedded support for learning everywhere, and transformed social interactions. Going big also requires thinking small: analyzing diagnostic micro-patterns to customize individual learning, sifting through millions of participants to find the ideal partners to aid each other’s growth. To reach massive with universal access and powerful outcomes, we must creatively expand our visions of platforms, pedagogy, and financing.
Chris Dede is the Timothy E. Wirth Professor in Learning Technologies at Harvard’s Graduate School of Education. His fields of scholarship include emerging technologies, policy, and leadership. His funded research includes grants from the National Science Foundation, the U.S. Department of Education’s Institute of Education Sciences, and the Gates Foundation to design and study immersive simulations, transformed social interactions, and online professional development. In 2007, he was honored by Harvard University as an outstanding teacher, and in 2011 he was named a Fellow of the American Educational Research Association. From 2001-2004, he was Chair of the HGSE department of Teaching and Learning.
Chris has served as a member of the National Academy of Sciences Committee on Foundations of Educational and Psychological Assessment and a member of the 2010 National Educational Technology Plan Technical Working Group. In 2013, he co-convened a NSF workshop on “new technology-based models of postsecondary learning.” His co-edited book, Scaling Up Success: Lessons Learned from Technology-based Educational Improvement, was published by Jossey-Bass in 2005. A second volume he edited, Online Professional Development for Teachers: Emerging Models and Methods, was published by the Harvard Education Press in 2006. His latest co-edited book, Digital Teaching Platforms: Customizing Classroom Learning for Each Student, was published by Teachers College Press in 2012.
Two tutorials will be offered in parallel and are free to registered attendees of Learning At Scale.
Building and Interpreting Adaptive Item Response Theory Models With Hands-On Training
- Eliana Feasley, Khan Academy
- Jace Kohlmeier, Khan Academy
- Jascha Sohl-Dickstein, Khan Academy and Stanford University
At Khan Academy we've delivered millions of tests to students to rapidly place them at their learning edge. Guacamole is the library we wrote and open sourced to train adaptive pretest parameters using Item Response Theory. We can rapidly learn multidimensional vectors of difficulty for hundreds of unique assessment items and millions of students at once, using Expectation Maximization and sampling. Guacamole also includes the engine we use to run adaptive tests. An essential part of learning at scale is learning complex representations of our students and problems rapidly from data. While it is true that the more data one uses, the more complex and detailed the model can be, anyone who has administered a battery of assessment items to a variety of users can easily use this tool both to learn about their data, and to more effectively deliver quizzes adaptively.
Guacamole is easy for non-programmers to use and powerful enough to support millions of students and hundreds of assessment items. This tutorial will consist of an extensive overview of Item Response Theory at scale, followed by an interactive session in which we will generate and interpret data. To prepare for the session, follow the instructions which will appear at http://khanacademy.org/r/guacamole before the conference begins.
Learning through Discussion: Foundations, Findings, and Future
- Carolyn Penstein Rose, Language Technologies Institute and Human-Computer Interaction Institute, School of Computer Science, Carnegie Mellon University
This tutorial will offer a review of the literature from the tutorial dialogue, classroom discourse, and collaborative learning communities on what is known about learning through discussion. We will begin with an overview of alternative theoretical frameworks for studying learning, including cognitive, sociocognitive, and sociocultural perspectives. Building on that foundation, we will explore existing frameworks from sociolinguistics, discourse analysis, and conversation analysis, and how they serve as lenses for viewing evidence of social and psychological processes at work in discussion interactions within learning contexts. We will view these frameworks through a computational lens, exploring how state-of-the-art language technologies and machine learning modeling techniques have begun to be used to operationalize them. The goal is for us as a community to review what we have already learned about what it takes for discussion to be beneficial for learning so we can consider what behaviors will be most valuable to measure and support and what strategic open questions we are in a position to address in new massive scale learning contexts.