A number of innovative projects at the Concord Consortium are helping students develop AI literacy, converse and collaborate with AI, and use AI as a learning and communications tool.

We’re helping students develop AI literacy and interest in AI-related careers. We’re partnering with the University of Florida, Texas Tech University, WestEd, and Florida Virtual School to develop a year-long AI supplemental certificate program, which will be integrated into high school math courses. We aim to provide high-needs students the opportunity to develop AI literacy and self-efficacy in solving problems using AI and learning AI topics, and to simultaneously improve their math learning and attitudes toward math. Virtual schools offer a flexible and accessible learning environment that can be accessed from anywhere with an internet connection, enabling students who may have previously struggled to access quality education to engage in rigorous academic programs.

We’re integrating language-based AI across the curriculum to create diverse pathways to AI-rich careers. Through ongoing collaboration among the Concord Consortium, Carnegie Mellon University, and North Carolina State University and networking in the AI education community, we have built a strong interdisciplinary team of AI developers and educators, STEM and humanities educators, learning scientists and designers, and experts on diversity, equity, and inclusion. The AI Education Across the Curriculum project extends this work by partnering with the San Joaquin County Office of Education in California and the Maryland Center for Computing Education. We’re working with school districts that serve student populations underrepresented and underserved in the field of AI. We’re refining our StoryQ app, which is designed for students in grades 6-12 to understand and apply machine learning with unstructured text data without coding, by developing a set of curriculum modules for students to explore language-based AI applications and related careers in their math, English language arts, and history classes.

We’re making data exploration accessible for blind and low-vision learners using AI. In partnership with Perkins Access Consulting at Perkins School for the Blind, and working closely with accessibility consultant Sina Bahram from Prime Access Consulting and AI consultant Vikram Kumaran, our Data By Voice project is addressing the critical need for accessible data science tools in K-12 education. We’re leveraging a large language model from generative AI technologies to create a multimodal data exploration environment for blind and low-vision (BLV) learners. Our AI-powered agent called DAVAI (Data Analysis through Voice and Artificial Intelligence) is embedded as a CODAP plugin. Designed to interpret BLV users’ verbal and typed commands, DAVAI provides an interface between the user, the generative AI model, and CODAP. By enabling BLV students to interact with data through voice commands, sonification, and AI-generated text descriptions, we aim to transform the educational experience and broaden participation in STEM.

We’re training machine learning to analyze students’ proportional reasoning. In partnership with Michigan State University, we’re taking the first steps to integrate AI into our CLUE (Collaborative Learning User Environment) platform with a problem-based mathematics curriculum. The Connected Mathematics Project emphasizes learning a variety of proportional reasoning strategies in 7th grade, and knowing when and how to apply them. Students need more immediate feedback and comparative examples of these strategies in reference to their own work. We are proposing AI evaluations to return this feedback promptly and consistently. To better train the AI model on students’ proportional reasoning, we developed student proportional reasoning arrows (SPArrows), which enable students to annotate relationships of proportionality across various mathematical representations.

SPArrows (curved arrows) in the CLUE platform.

Human coders can look at students’ use of SPArrows to more accurately tag their work against a proportional reasoning rubric. This tagging will be used to train a machine learning model that can evaluate student work on the rubric. We are currently exploring 1) how AI systems can group student work based on similar mathematical problem-solving approaches and suggest strategies that may not come up in class discussions and 2) how to track and report on individual student learning over time. (Learn more in this article in Digital Experiences in Mathematics Education.)

We’re supporting secondary students’ front-end engineering design skills. The Mobile Online Studio project with the University at Buffalo and the University of Michigan is engaging middle and high school students in front-end design activities in Earth science and engineering. Front-end design—which includes defining and exploring problems; gathering information about technical, social, and contextual elements; developing constraints; and generating diverse ideas—presents a unique opportunity for student-driven integration of science and engineering with social factors to generate interest in engineering and STEM and computational thinking competencies. We’re incorporating an AI-powered virtual design mentor into the CLUE platform to suggest ways for students to generate and explore diverse design ideas as they tackle complex challenges in socioscientific reasoning and design thinking by collaboratively engaging in front-end work. We hope to cultivate a sense of community and civic engagement along with critical thinking skills about how to solve problems creatively.

We’re developing an innovative curriculum and datathon to teach students about bias in AI and medical databases. While future STEM careers are becoming increasingly reliant on AI and data science, bias-related limitations within AI and within medical databases can create inequities in society. A new semester-long course teaches high school students how embedded bias in machine learning affects healthcare discussions and decisions. At the end of the course, teams of at least two high school students, one STEM teacher, an undergraduate computer science student, a clinician, and a data scientist collaborate to solve a problem in a two-day datathon. The goal of the Data Science, AI, and You in Healthcare project is to foster community connections between educators, researchers, clinicians, and local stakeholders and prepare underrepresented students for STEM jobs—ultimately in fields where their background and experience can help mitigate bias embedded within AI healthcare models and improve outcomes for all.