Collaborative Human-AI Learning Ecologies for STEM+C Lab

The CoHALE Lab designs and studies collaborative human-AI learning ecologies for STEM and computing. We build learning environments in which students, teachers, and AI agents work on disciplinary problems together, from sensor-based inquiry in school gardens to generative-AI partners in teacher learning and XR maker experiences in teacher preparation. We study these environments in classrooms, courses, and professional learning with partner teachers and investigate how the designed roles of AI participants shape reasoning, learning, collaboration, and agency.

Three research strands

Disciplinary computational practicesHow computational thinking, data practices, and AI literacy are learned, enacted, and assessed within disciplines.

High-quality collaborationHow collaborative competence develops, what models of collaborative quality capture or miss, and how participation and trust shape joint work.

AI agents as participants in learning ecologies How the designed roles of AI agents, including simulation partners, cognitive partners, and peer mentors, shape reasoning, learning, and agency.

How we work

Co-design with teachers

Partner teachers co-design the units and modules they will teach and take increasing ownership of them as a project matures.

Design-based research in real-life settings

Iterative cycles in classrooms, courses, and professional learning, using mixed methods including surveys, student and teacher artifacts, classroom video, AI interaction logs, and interviews.

Theory in design and analysis

We operationalize theoretical frameworks through learning designs, scaffolds, rubrics, matrices, AI agents, and analytic tools. Evidence from participants' activity helps us examine what these frameworks capture or miss and refine both the designs and theoretical accounts of practice.

Projects

NSF CAREER2026–2031 · PI

3AforCT. Sensor-based school-garden inquiry in two rural Texas districts, carrying data practices across the inquiry cycle; an AI peer mentor that models expert reasoning and fades; teachers who co-design the unit they teach. Built on two pilot grants (2024–2026): AIM4Rural (Emerge, $49K) and AI-Enhanced CT & Agricultural Science (Catapult, $30K).

NSF IUSE2026–2031 · Co-PI

AI-Driven Professional Learning. A generative-AI cognitive partner that scaffolds preservice teachers through creativity-directed mathematics tasks, implemented across eight teacher preparation programs.

Design studiesongoing

Computing and AI in teacher preparation. XR maker experiences and AI-literacy modules through which preservice teachers take up computational and AI practices inside their own subjects, working with and supported by the technologies they will teach with. Recent work: a diagnostic framework for reading makerspace process data as evidence of computational thinking (Interactive Learning Environments, 2026); safeguards and professional learning needs for AI integration (Cogent Education, 2026).

CSCL lineongoing

Collaborative competence and the technologies that support it. Whether teams can recognize and improve their own joint work – from a system that returned to student groups the criteria their discussions were judged by, to a chat-based simulation partner for practicing collective sensemaking – and what models of collaborative quality miss when the talk turns to identity and difference.

Join the lab

Interested in how students and teachers learn with AI, in computational thinking inside STEM subjects, or in collaboration research? Graduate students and collaborators work across the projects above with partner teachers and schools. Prospective students and collaborators are welcome to reach out.

Email Dr. Aldemir