Published: Jan. 6, 2022

During this quarter, our Strand 3 researchers continued working directly with K–12 students and teachers to ensure iSAT fulfills its mission to create equitable and socio-collaborative learning experiences for all students. Strand 3’s work over the quarter amplified the voices of students from all backgrounds to inform the data analysis work of our Strand 1 researchers and the team science and human computer interaction work of our Strand 2 researchers.

Check out this video titled How Artificial Intelligence is Making DPS Stem Classrooms More Equitable by Denver Public Schools to see students, teachers, and our researchers in action! 

Helping Students Understand AI

Strand 3’s Co-design team engaged with K–12 students and teachers to co-design a middle school unit focused on helping students understand, critique, and design basic AI systems. The educators who have been involved include two leaders from iSAT’s partner districts, as well as teachers of educational technology, English Language Arts, World Geography, Mathematics, and history. They also had an educator from an after-school education program focused on Media Literary and Agency.

The goal is to complete development of the unit in time for a small group of 3-4 teachers to implement the unit in spring 2022, before iterating on the unit in a way that integrates the work of Strand 1 and 2. Strand 1 and 2 researchers are already part of the extended co-design team. Since September, the team has completed a storyline for the unit, that is, a detailed outline of the lesson level questions students will investigate, what they will do in each lesson, what they learn, and what questions emerge from each lesson. In addition, the team has drafted roughly half of the lessons for teachers and engaged teacher co-designers in a weekend session where they solicited feedback on initial lesson designs.

In addition to continuing to design the unit, the team, led by postdoctoral scholar Dr. Areej Mawasi, conducted interviews with co-design participants to learn more about their experience of co-design.A key finding from the analysis was that the co-design process allowed for collaborative learning and interactivity among participants with different kinds of expertise and allowed for intergenerational collaboration. They also found intentional design aimed toward bringing together diverse forms of expertise was perceived as a “safe space” for collaboration. Educators and youth who took part in the co design process this summer were positive about the importance of this work, which, as one educator noted, “. . . disrupts power dynamics. It disrupts our notion of what curriculum can be and look like. And then I  think ultimately it infects the K-12 space with something that actually seats or has the potential to seat young people as co-conspirators in developing their own learning.”

Creating Immersive AI Experiences for Students

While the Co-design team worked toward helping students understand how bias can occur in AI applications, the Sensor Immersion team implemented a student-centered curriculum co-designed with teachers to immerse K–12 students in the fundamental concepts of AI.

SVVSD students work on a Sensor Immersion task.

Strand 3 researcher Quinten Biddy helps SVVSD students with a Sensor Immersion task.

In the Sensor Immersion curriculum, students investigate a programmable sensor system called the “DaSH” (Data Sensor Hub) to engage in computational thinking. The DaSH allows students to design, program, and build a custom system to help them gather data about the world around them. They then use this data to answer questions they have and create solutions for real world problems. For example, students worked with a sensor that measured how much water was present in the soil of a plant to determine how much water was needed to help the plant survive.

The Sensor Immersion curriculum was implemented by 3 St. Vrain Valley School teachers and 9 Denver Public School teachers, reaching about 250 students in St. Vrain Valley School District (SVVSD) and 675 students in Denver Public Schools (DPS). The team also analyzed data from the pre/post surveys of student experiences of lessons 2 and 4 of the iSAT Sensor Immersion unit and started the process of prototyping visualizations, report, and dashboard  content for teachers and researchers to use. To help collect this data, the team worked with a group of CU Boulder Senior Software Engineering Capstone project students to instrument the Makecode environment to collect a record of all actions performed by the students. They also worked with the cross-strand iSAT Infrastructure team to build and test iSAT’s new audio/video recorder.

In the coming quarters, the team would like to build an extension unit with St. Vrain Valley School teachers using smart cities as a context and incorporating more AI learning standards from the AI4K12 frame-work. The team also hopes to build asynchronous professional learning opportunities related to Sensor Immersion to improve scalability.

SVVSD students work on a Sensor Immersion lesson.

iSAT K—12 Partner / Strand 3 researcher Axel Reitzig looks on as SVVSD students work on a Sensor Immersion lesson.

What Should our AI Partner Do — and Not Do?

This fall the Learning Futures team processed the data and interactions from the summer’s Learning Futures Workshop. This workshop engaged 30 high school-aged youth from California and Colorado to deliberate the role of AI in supporting classroom collaboration. During this interactive 5-day summer workshop, these youth expressed and explored their hopes and concerns of the use of AI in the classroom. The team focused on the youths’ dreams for the best-case scenario use of AI in classroom settings, challenging the students to expand their imaginations about the possible ways AI could sup-
port their learning. The team also focused on the youths’ concerns with having an AI Partner in their classroom and asked them to think about what they wouldn’t want the Partner to be able to do. 

The researchers found that the youth consistently expressed their desire for affirming interactions from an AI Partner. They want their contributions heard, attributed, and to have meaning within the classroom. However, they did not want the Partner to report any poor behavior to their teacher. The youths also want agency over the AI Partner and would like the ability to turn off the Partner at their will, but they felt they would be willing to “give up” data about themselves in exchange for features they value.

The team is currently using this information on the dreams and concerns of youth to inform the larger Institute on the role of our AI Partner in the classroom. The team hopes to have youths identify and analyze system inequalities at school and within institutions for their next workshop.