The idea for StudyBuds, an early-stage AI practice product, was rooted in my belief that students are more likely to retain what they learn when they actively encode it during practice, something grounded in learning sciences research. One of the most powerful methods to practice is “learning by teachers”, where students deepen their understanding by explaining what they know to others.

With StudyBuds, we built an environment where students could teach AI partners (their StudyBuds). These teachable agents challenge students to confront misconceptions, provide and explain evidence, and engage more deeply with content. In fall 2024, we tested StudyBuds with a small group of secondary social studies and science teachers in Tennessee and Texas. Teachers shared really encouraging qualitative feedback – students were experiencing productive struggle and felt motivated to work through that struggle alongside their StudyBuds.

In many organizations, this type of early signal would be enough to move quickly toward commercializing StudyBuds. At Teaching Lab Studio, however, our main focus is to provide evidence of efficacy – ensuring that a tool truly benefits students before scaling.

Any developer of an early-stage edtech tool knows how difficult testing is. The hurdles around forging school partnerships, collecting student data, and getting whitelisted by local IT are significant – and for good reason! Schools are doing their due-diligence to ensure strong, secure products are going in front of students. Despite any hurdles, this stage is critical; we need to know if what we’re building makes a real, positive impact. 

This is why we at StudyBuds partnered with Leanlab Education, an independent nonprofit research organization. Leanlab helped us connect with schools that were already prepared for research partnerships, while providing an impartial, third-party to prevent any unintended bias from slipping into the process or results. 

In spring 2025, Leanlab sourced ten science and social studies teachers (grades 5-10) from nationwide public, private, and charter schools to answer the following key questions: 

1. Does StudyBuds positively impact student learning?

2. Is StudyBuds practical and valuable for real classrooms?

More than 600 students participated in this study. Here’s how the study worked: 

• First, students in each classroom took a pre-assessment (an exit ticket) before using StudyBuds.

• Then, teachers then implemented three StudyBuds modules aligned to their scope and sequence over a period of two to three weeks. They had students complete additional exit tickets. 

• Additionally, teachers completed mid-point surveys and joined end-of-study focus groups to share feedback on usability, barriers to implementation, and perceived value as it relates to student engagement and productive practice.

  
  

Leanlab reviewed teacher feedback and compared students' exit ticket performance before and after using StudyBuds, and we saw exciting results!

The quantitative data showed a statistically significant improvement in student scores over time (p = .03). Across the 627 participating students, average scores rose consistently from pre- to post-test. In some classrooms, average scores rose by more than 20 percentage points after using StudyBuds modules! Growth was observed across all grades (grades 5-10), suggesting value at both middle and high school. While this was not a quasi-experimental study (no control group), the findings offer strong early evidence that StudyBuds can positively impact learning.

In addition to the positive student scores, teacher feedback added incredibly valuable context. Educators highlighted StudyBuds’ role in boosting student engagement and motivation, their ability to provide immediate feedback, and for driving deeper thinking. Teachers also offered constructive feedback on their experience, some of which we have already implemented (like simplifying the start-up process) and others that we are actively working on (such as providing teachers with better real-time visualization of student data).  

Data-driven, early-stage efficacy validation is possible, and StudyBuds is living proof. This study represents a pivotal moment in our product journey. We’ve always believed in learning science – that rigorous practice leads to better retention. Now, we have evidence to support that belief, and we will continue to iterate towards our mission of making practice more productive and feelings of self-efficacy for students. 

Just like students, we’re learning and growing every day. Reach out to me at adam.franklin@teachinglabstudio.com, and explore StudyBuds for free at studybuds.io.