Creating AI-enabled Collaborative Learning Environments: Strand 2
Our team science experts in Strand 2 strive to better understand how students, AI, and teachers can collaborate effectively in both classrooms and remote learning contexts. Their work during our first year focused on designing our AI-enabled Collaborative Learning Environments (AICL), which are set to enter the lab testing phase this fall.
A schematic showing the small group, in-person setup for the AICL.
These environments are the hardware/software interface for the AI Partners our institute is developing. They serve as a research tool, data collection tool, a tool for AI model training and testing, and a learning tool. The AICL merges the foundational AI achievements from Strand 1 with the use-inspired classroom experiences and curriculum developed by Strand 3 to help Strand 2 develop this AI support for classrooms. Furthermore, the AICL has been designed with an adherence to the principles put forth in the Responsible Innovation Framework. Not only does the AICL serve as the underlying technology of the AI Partner, it also provides a tangible platform to unite research conducted by the different strands鈥攁 critical goal to avoid the 鈥渙ne-off鈥 studies achieved in academic silos.
The team includes one of our development partners, Curve10, who are leading the development of WoZWare, the software interface that will enable our researchers to test and iterate the different roles and functions of our AI Partner. What started as a collection of ideas in our second quarter and a wireframe in our third quarter, WoZWare is now ready for our researchers to playtest and continue to develop as we wrap up our fourth quarter and head into year two.
To that end, research in the fourth quarter included drafting and receiving approval for an IRB protocol for CU 麻豆影院 data collections with the AICL. We also selected and purchased the hardware equipment for the AICL experiments, including Chromebooks, a Yeti Microphone, a Pal Mini Vu camera, and Microsoft Kinect cameras. Pilot testing was conducted on the modified CPS version of the Sensor Immersion Unit (creating an automated plant watering system through team programming), and results of the pilot studies were used to make minor modifications to the stimulus materials.
Data collected from Eagle County students鈥 work with iSAT researchers during STEM week will help inform Strand 2鈥檚 AICL work.
To create an effective and engaging AI Partner that helps teachers better reach their students, our Strand 2 researchers spent time this quarter synthesizing research on classroom orchestration. The team also conducted interviews with teachers, and participatory design sessions in which teachers worked on ideas for their dashboard to create a mock dashboard focused on facilitating small group classroom discussions. Other Strand 2 members are also researching key creative problem solving skills that our AI Partner will want to help cultivate in future classrooms.
Lastly, the end of the fourth quarter saw the creation of a new working group in Strand 2 focused on the 鈥榥euroscience of student AI teaming鈥. The new working group began to hold meetings and conceptualize research questions and hypotheses to address in year two and beyond, using an extension of the AICL lab environments.