Machine learning (ML) models are fundamentally shaped by data, and building inclusive ML systems requires significant considerations around how to design representative datasets. Yet, few novice-oriented ML modeling tools are designed to foster hands-on learning of dataset design practices, including how to design for data diversity and inspect for data quality. To this end, we outline a set of four data design practices (DDPs) for designing inclusive ML models and share how we designed a tablet-based application called Co-ML to foster learning of DDPs through a collaborative ML model building experience. With Co-ML, beginners can build image classifiers through a distributed experience where data is synchronized across multiple devices, enabling multiple users to iteratively refine ML datasets in discussion and coordination with their peers. We deployed Co-ML in a 2-week-long educational AIML Summer Camp, where youth ages 13–18 worked in groups to build custom ML-powered mobile applications. Our analysis reveals how multi-user model building with Co-ML, in the context of student-driven projects created during the summer camp, supported development of DDPs including incorporating data diversity, evaluating model performance, and inspecting for data quality. Additionally, we found that students’ attempts to improve model performance often prioritized learnability over class balance. Through this work, we highlight how the combination of collaboration, model testing interfaces, and student-driven projects can empower learners to actively engage in exploring the role of data in ML systems.
7 November 2025 in Early Childhood Education Journal
Artificial intelligence (AI)-powered tools are becoming increasingly ubiquitous in children’s environments. However, three issues remain to be addressed: (1) an “AI divide” in access and use of AI for learning between socioeconomically advantaged and disadvantaged children, (2) accessibility and opportunity gaps in AI literacy acquisition between these two groups, and (3) an ethical gap in AI use by children. In this review- and policy-based article, we first synthesize some current research on AI applications in early childhood education (ECE), with a focus on equity, AI literacy, and ethical concerns. We then provide actionable policy recommendations for education leaders and practical guidance for teachers. Specifically, we encourage education leaders to apply an equity lens by affording children and teachers AI-related resources and opportunities to develop AI literacy. This affordance can, in turn, empower teachers to engage children in leveraging developmentally appropriate AI tools for exploration and learning. Furthermore, we recommend that education leaders invest in AI infrastructure for teachers as well as children (and their families) that provides clear guidelines in promoting ethical, safe, and responsible use of AI. In particular, we advocate focusing on five key priority areas — AI infrastructure, teacher capacity, teaching sensitivity, AI safety and security guardrails, and strong teacher-parent partnerships.
11 November 2022 in Investigations in Mathematics Learning
A common approach to scaling up a professional development program is for the researchers who designed the program to prepare teacher leaders to facilitate it at their schools. When researchers eventually leave, however, teacher leaders may receive less support. To ensure that teacher leaders continue receiving support, researchers can prepare district mathematics specialists to assume responsibility for preparing the teacher leaders. Little is known, however, about district mathematics specialists’ role in sustaining, and potentially adapting, professional development programs. We examined district mathematics specialists’ facilitation of an adaptive teacher leadership preparation program. Program sessions were originally facilitated by researchers then by the specialists. We analyzed the adaptations specialists made to the sessions over four years and the rationales underlying these adaptations. Specialists maintained the program’s overall structure, continuing to model the facilitation of core program activities that teacher leaders would then facilitate in their site-based professional development workshops. However, they modified the thematic focus of these activities to address district goals, interests, and priorities. Adaptations were informed by specialists’ intimate knowledge of what was occurring in district schools. This approach maintained activities supportive of teacher learning, but also demonstrated that the specialists took increasing ownership over the program by adapting it.
11 June 2025 in The Journal of Educational Research