Empowering Educators in AI: Insights from Co-Designing an AI Microcredential with and for K-12 Educators

Nicole M. Hutchins, Shan Zhang, Joanne R. Barrett, Maya IsraelAAAI 2025 · DOI

Introduction

As Artificial Intelligence (AI) continues to reshape the educational landscape, K-12 teachers often find themselves at a crossroads. While the potential for AI to enhance learning is vast, many educators feel ill-equipped to integrate these tools effectively or ethically. Existing professional development (PD) options frequently miss the mark, offering overly theoretical content or acting as marketing vehicles for specific products rather than providing practical, classroom-ready training.

This paper details a co-design process between researchers and K-12 educators aimed at building a foundational AI microcredential. By involving teachers directly in the design phase, the authors identified critical requirements for AI PD—such as the need for "unfiltered" insights into AI limitations and "just-in-time" resources—that bridge the gap between theoretical AI potential and the practical constraints of modern classrooms.

Methodology

The researchers employed a participatory co-design approach involving six stakeholders: four middle school teachers and two instructional facilitators. This methodology focused on "precedent knowledge"—understanding the unique classroom realities defined by district regulations, state standards, and significant time constraints.

Key qualitative techniques included:

  • Conjecture Mapping: A framework used to link high-level design ideas to specific classroom outcomes through "mediating processes."
  • Teacher Superpowers Card Sorting: An activity where educators ranked professional learning goals from "most important" to "least important" to identify their primary needs.
  • Inductive Coding and Memoing: Transcripts from co-design sessions were analyzed to identify recurring themes and requirements.
Educators ranked professional learning goals to identify primary needs during the "Teacher Superpowers" card-sorting activity.

Key Insights: Bridging the Gap

The co-design process revealed several "aha moments" that challenge standard assumptions about teacher training:

Agency within Constraints

While teachers value the agency to choose AI tools, they are heavily restricted by district-level limitations and state testing requirements. Effective PD must acknowledge these "handcuffs" rather than ignoring them, providing strategies that work within existing regulatory frameworks.

The Demand for "Unfiltered" Knowledge

A striking finding was the educators' desire for transparency. Teachers expressed a strong need to see where AI fails. They want to understand the limitations, biases, and ethical pitfalls of AI to protect their students and maintain an objective perspective, rather than receiving a polished "sales pitch" for the technology.

Modernization on Teacher Terms

Teachers are not looking to replace their existing pedagogy with a total digital overhaul. Instead, they seek AI resources that support current practices—including paper-based activities—allowing for a gradual and relevant modernization of the classroom.

Just-in-Time Learning

Traditional, lengthy summer PD sessions are often insufficient for the fast-paced school year. Educators expressed a preference for "short and effective" (less than 5-minute) asynchronous videos and resources they can access "just-in-time" to solve immediate classroom challenges.

Results: The Updated Conjecture Map

The primary output of this research is an Updated Conjecture Map that incorporates specific teacher-requested features. This map serves as a blueprint for the microcredential, ensuring that the "Embodiment" (tools and tasks) leads to meaningful "Mediating Processes" and, ultimately, the desired "Outcomes."

The Updated Conjecture Map incorporates teacher-requested features such as "short" experiential applications and a focus on district-level limitations.

Key refinements in the updated map include:

  • AI Curriculum: Inclusion of "Short" experiential applications and resources designed to be "Taken to the Classroom" immediately.
  • Mediating Processes: A new focus on "district-level limitations" and "behavioral changes" in students, such as identifying over-dependence on AI tutors.
  • Outcomes: Expanded goals including "Creating an unfiltered, unbiased opinion about AI" and "making modernization changes... when applicable."

The authors also developed two specific case-study modules: Teachable Machine for exploring data bias, and AI Critical Eye for identifying ethical concerns like data privacy and "AI dependency."

Limitations and Future Directions

While the co-design process provided deep insights, the study is limited by its small sample size of six participants, which may not represent the full diversity of K-12 environments. Furthermore, while the design of the microcredential is well-documented, large-scale empirical evidence of its effectiveness on teacher and student outcomes is still pending.

Future work will involve refining the modules based on initial feedback—such as simplifying data management tasks—and scaling the microcredential to a broader audience of elementary and secondary educators to validate the co-designed framework across diverse educational contexts.

Summary updated

BibTeX
@inproceedings{hutchins2025empowering,
  author = {Nicole M. Hutchins and Shan Zhang and Joanne R. Barrett and Maya Israel},
  title = {{Empowering Educators in AI: Insights from Co-Designing an AI Microcredential with and for K-12 Educators}},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year = {2025},
  doi = {10.1609/aaai.v39i28.35186},
}
Figure