Â鶹ӰԺ

AÌýbig part of iSAT’s mission is to create an AIÌýPartner who helps teachers facilitate collaborative and equitable learning outcomes forÌýsmall groups in the classroom. This quarter,Ìýthe team science and human-computer interactionÌýexperts in Strand 2 have made important progress inÌýunderstanding the framework needed to help smallÌýgroups develop successful collaboration skills andÌýhow to support teachers help their students developÌýthese skills.

What Skills Should Our AI PartnerÌýDevelop?

Strand 2’s Framework Development team is dedicated to figuring out a huge overarching questionÌýfor our institute: Exactly what skills should our AIÌýPartner help students develop?Ìý

To answer this question,Ìýthe team hammered awayÌýat developing a framework of collaborative problem-solving (CPS) skills our future AI Partner willÌýlearn so it can assess and encourage small groupÌýinteractions.ÌýThe team’s main goals are to determine behaviorsÌýand skills that underlie successful collaborative problem-solving. The team formulates metrics to measureÌýthese skills with a focus on how skills emerge overÌýthe course of problem solving and with the goal toÌýmodel metrics in real time, such as the state of CPSÌýcan be fed to the teacher and/or AI Partner.

While closely collaborating with Strand 1’s Reinforcement Learning team on the CPS skill measurement, the team will use data collected as part of theÌýAI Collaborative Learning (AICL) environments workÌýunderway by Strand 2 to test the assumptions ofÌýthe framework and further develop metrics used toÌýasses CPS. In addition, the team is also using dataÌýcollected from a previous NSF-funded project onÌýremote CPS, in which teams collaborate within aÌýMinecraft (a popular sandbox video game) task, toÌýcreate influence metrics that will be adapted as partÌýof this framework. This study is the first step in formulating a finalized metric for the equitable interaction skill found in the framework. The team alsoÌýcollaborates with Strand 3 to assess what actionsÌýshould be taken in response to the measurement ofÌýaÌýskill.

How Will Our AI Partner Talk toÌýTeachers?

While the Framework Development team has beenÌýhard at work figuring out the most effective CPSÌýskills to measure and implement, the ClassroomÌýOrchestration group has spent the past quarterÌýcollaborating with K–12 teachers on the best ways toÌýpresent classroom data to teachers in real time.ÌýThe Classroom Orchestration team conducted twoÌýparticipatory design studies with eight WisconsinÌýarea teachers to determine what type of informationÌýwould be most useful to them on a teacher dash-Ìýboard. With the teachers’ input, the team createdÌýdashboard mockups and further iterated on theÌýdesigns with a focus on specific features.

Initial analysis from the studies showed that teachers emphasized that their needs change based onÌýthe goals of each specific class. For example, someÌýteachers preferred a general overview that directsÌýthem to where they need to focus their attentionÌýwhile others were more interested in diving deeperÌýinto the data to see classroom trends, to improveÌýcurrent and future instruction. While teachers foundÌýitÌýhelpful for group and class level information to beÌýpresented during class, most preferred information atÌýthe individual level for post-class analysis.Ìý

Based on the dashboard designs created throughÌýthe Wisconsin-area participatory design studies, theÌýteam began mock-ups for dashboards to use in theÌýÂ鶹ӰԺ iSAT Lab. TheÌýiSATÌýLab dashboard mockups are being designed with theÌýSensor Immersion work by Strand 3 in mind, but theyÌýare mostly focused on designs that can span a range ofÌýcurriculum activities. In the coming quarters, the teamÌýplans to develop and use the first version of theirÌýdashboard with six teachers in Wisconsin.

iSAT Lab Experiments

Strand 2 also led the effort to create iSAT’s very ownÌýlab in CU Â鶹ӰԺ’s Center for Creativity and Innovation (CINC) in Â鶹ӰԺ, Colorado. The lab functionsÌýas both a research study facility and a space forÌýallÌýstudents to gather, network, and collaborate.

iSAT researchers developed studies for this lab in which students, staff, and faculty at CU Â鶹ӰԺ workÌýon collaborative problem solving while our researchersÌýcollect video and audio recordings of theirÌýinteractionsÌýon the same equipment used by our researchers toÌýcollect classroom data. The main goal for this lab workÌýis to test collected data on Collaborative ProblemÌýSolving skills to advance teamÌýscience and to look intoÌýhuman-AI teaming strategies. These recordings willÌýalso further help our Strand 1 researchers with theirÌýdata annotation, multi-modal, and content analysisÌýefforts by providing themÌýwith cleaner, easier-to-process data than what is collected in noisy K–12 environments. The various teams will then use their improvedÌýdata analysis processes to optimize our K–12 classroom data.