Growth Mindset Activity Maker Prototype
Designed with teachers, using Microsoft’s Azure AI® Platform
Backdrop
In 2023, Committee for Children was awarded a grant from the Gates Foundation to invest in innovative projects leveraging AI to enhance equitable mathematics outcomes, particularly focused on supporting middle school Black, Latino, and low-income students.
Committee for Children (CFC), one of the largest providers of Social-Emotional Learning (SEL) programs knows very well that classroom climate greatly impacts the learning conditions. Educators have a tall order to fill, to support an environment that meets all their students' needs in addition to teaching math.
While Committee for Children is not an expert in math or artificial intelligence (AI) our team was excited to learn what might be possible with the intersection of SEL, math challenges, and AI. Can AI help integrate custom SEL opportunities in the classroom while supporting math learning objectives?
As the lead UX researcher and designer on the project with an acute awareness of the potential pitfalls of AI development and deployment, I decided that this project must deeply partner with educators and be conscientious of where AI does and does not best respond to student needs.
The Team
• Product Manager
• Researcher and Designer (myself)
• HCD advisor from Oxford University
• 2 engineers
Discovery Phase Highlights
Check our Mindset
The product team had an initial design solution before conducting any research. This raised a red flag for me, more inquiry is needed before jumping to conclusions. After discussion, my manager and I proposed to the team to shift our mindset from “solution-minded” to "learning-minded".
How might we maintain the stance of a learner who is not trying to “solve” anything?
Plan Together
Given the short window of time for the grant, I proposed a rough project plan, inpired by the double-diamond approach, to accomplish our goals and reach key milestones. We revised it together as a full team which helped us move forward and use each precious week wisely and intentionally.
Empathy First
What is it like to be a middle school math teacher, supporting ~25 different students, each with a variety of needs, confidence levels, cultures, and degrees of home support?
After reading several journals and articles on how math-specific anxiety affects the learner’s experience, I wanted to spend time with Middle School Math teachers directly. I hosted two focus groups to learn more:
How do middle school math teachers assess when a student needs help in math?
How confident do they feel in their assessment?
Do educators link SEL skills as a means of supporting students? In which ways?
What stories do educators have when responding to and supporting students?
What affinity do these teachers have for using new technology? Specifically AI? What are their hopes and doubts?
The focus groups were rich, interactive, and inspiring. Educators shared a lot with us about the root causes of why students lose motivation in math, and how teachers can or cannot help.
Here are a few takeaways and learnings from the focus groups:
The focus groups identified barriers and root causes for students of color and low-income families as often the conditions outside of school, this includes the unavailability of family support, additional responsibilities at home, and reduced resources for tutoring and supplementary help. This can compound over time and significantly reduce motivation and self-confidence.
A common pain point for middle school math teachers is for students to improve their self-confidence. Teachers long for kids to arrive in the classroom ready to learn and at least be willing to try, even if they don’t know how, or won’t get it right.
Teachers are interested in AI assistive technology to support them, and their students. They shared concerns as well. They expected that a meaningful tool should:
Provide relevant and adapted suggestions,
Have an awareness of classroom context (class size, social dynamics, grade, subject)
Integrate SEL and Math problems
Be feasible to implement (time to complete and transition, prep needed)
Be trusted and is designed from quality pedagogical resources
Not replace collaboration with colleagues
Have consistent and quality results
"I want kids to have more confidence in themselves and the skills that they have. The kids automatically think they can’t do something, but they forget all the tools they do have. How can I help the kids see and feel that they are successful, even in a new more challenging problem?" - Math Educator
Below is an example of an activity in our focus group. Educators brainstormed causes of motivation loss in students universally and through an equity lens. We then looked at possible correlations of how SEL may address the motivational barriers.
Async Research Community (ARC) Highlights
Design and host a participatory design process
With the support of an HCD advisor from Oxford University, I created an invitation for middle school math teachers to participate in a 5-week research community engagement. We called this an Asynchronous Research Community (ARC) and used Slack and email to learn together about weekly research questions.
The ARC method allows:
Participation in as much or as little as they're interested in.
Flexible compensation and scheduling.
Activities to happen throughout the week in the classroom.
Collaboration, engagement, and support with other educators.
Participants to help guide the research process. Each activity I designed was informed and reflected by the previous week's learning outcomes.
Primary tools: Slack, Miro, Email
The activities
Diary study of observations and situations happening in the classroom
Themes mapping
Identification of intervention opportunities using iceberg model map
Brainstorm of low-fidelity prompt design concepts
“Wizard of Oz” testing of low-fidelity design outputs
Refine AI model, prototype AI-assistant with engineering
Test activities in the classroom in real-time
Exit survey and reflection
Sample activity: Mapping Root Causes of Fixed Mindset
Notice and reflect together. What patterns and possibilities are present?
What themes or patterns do you notice in this iceberg map of stories?
Did this spark any full classroom ideas of intervention opportunities to support a growth mindset?
What kind of classroom support do you wish was possible to access, even without knowing how?
Sample activity: Design an activity maker prompt flow
Growth Mindset Activity Maker AI Chat Prototype | Try it
Outcomes
After our research and design engagement, we had a fully-functioning AI assistant that generated student-relevant SEL-informed math learning activities.
The AI assistant criterion was directly informed from educators.
Is this activity relevant to my students' interests?
Is this activity helping re-teach or reinforce concepts that students may have missed?
Is this activity helping to strengthen a growth mindset?
Is this activity modifiable to better fit my classroom’s needs or learning objective?
In November 2024, my colleague and I will be presenting about this project at the Annual Conference for Middle-Level Education. We also heard some very encouraging comments from the ARC after the project finished.
Participant Closing Reflections
"I am hopeful that it makes teachers' lives easier. I am hoping that is something that is accessible to ALL teachers in ALL districts."
"I feel like it would help raise engagement, give them confidence boosters, and bring in a fun way to practice math."
"This group was a great opportunity for me to connect with others in similar situations and see how they have helped their students. I am a better math teacher because of this project. Thank you for this opportunity!"
"I appreciated hearing ideas and struggles from other teachers. It was nice to know the issues I have are not isolated to my school or classroom."
Closing Reflections
It was a personal joy to be on a learning journey with educators in the new terrain of AI. I was intimidated at first to approach a project where there are so many unknowns, but with the support of the Gates Foundation, our advisors at Oxford, and the CFC team, I found the project to be incredibly rewarding and encouraging.
If we were to continue, we would need to conduct more UX research and design. We could create a UI, do stringent instructional design reviews and toning of the AI content, and do more testing to get a sense of student interest and resonance. But given the time constraints, I still believe we accomplished and learned a lot together. I am curious about the feedback at our presentation at the Middle School conference in November 2024!
Special Note
Thank you to the dedicated team of engineers, advisors, management, researchers, product engagement specialists, and instructional designers. Their collaborative efforts were the backbone of this initiative. I am deeply grateful for the generous support from the Gates Grant, which was instrumental in facilitating key collaborations, notably with the Allen Institute and our co-participants such as the EBR School District.
This remarkable project would not have been possible without the 23 middle school math teachers who contributed their expertise, candor, and wholehearted teaching experience. Their insights were pivotal in guiding us on a promising path to an intervention to support equitable student thriving.