Research map / Image generation / Text-to-image diffusion
A Picture Is Worth a Thousand Words: Co-designing Text-to-Image Generation Learning Materials for K-12 with Educators
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Designs and pilots a 6-hour K-12 teacher professional development curriculum on text-to-image generation (technical, applied, ethical), co-designing classroom prototypes with 30 educators across two workshops.
Problem
K-12 educators need to understand text-to-image generation tools (technical workings, classroom uses, ethical implications) to prepare students, but few teacher professional development materials cover generative AI of this kind.
Approach
A six-module, hands-on teacher PD resource on text-to-image generation grounded in adult learning principles and communities of practice, where educators create TTIG tool artifacts and then co-design K-12 learning materials, with ethics woven through every module.
Key results
- Thirty teachers across two workshops (12 and 18 consenting) completed the 6-hour materials; educators designed 19 K-12 learning prototypes.
- Teachers' self-reported confidence with generative AI rose from pre-survey (n=15) to post-survey (n=10) for one illustrated prototype's audience, per Figure 2 trends.
- Participants identified biases in generated portraits and imbalanced training data as causes of unequal outputs.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use the TTIG educator learning materials and co-design workshop | K-12 educators · Teacher professional development workshops | self-reported confidence using and understanding generative AI | Helped |
Surprising findings
- Self-portrait activity was replaced in workshop 2 because discussion showed it could harm K-12 learners from race, ethnicity and gender minorities.
Limitations
- Small sample with incomplete consent-based data (12 and 18 participants) and small pre/post survey counts.
- Evaluation relies on self-reported confidence and qualitative observations rather than validated learning assessments.
Experimental setup
Datasets: LAION-5B
Metrics: pre/post survey self-reported familiarity and confidence learning outcome attainment observed in workshop activities number of teacher-designed prototypes
Builds on: communities of practice adult learning theory Big 5 Ideas in AI
Closest papers by meaning
- Constructing Dreams Using Generative AI (AAAI 2024 · similarity 0.85)
- ImageSTEAM: Teacher Professional Development for Integrating Visual Computing into Middle School Lessons (AAAI 2024 · similarity 0.84)
- Word2Vec4Kids: Interactive Challenges to Introduce Middle School Students to Word Embeddings (AAAI 2025 · similarity 0.83)
- From Consumers to Critical Users: Prompty, an AI Literacy Tool for High School Students (AAAI 2024 · similarity 0.83)
- Advancing Research on Equitable AI Education Through a Focus on Implementation: Insights from a Middle School Computer Vision Module Beta-Test (AAAI 2025 · similarity 0.83)
- “From Unseen Needs to Classroom Solutions”: Exploring AI Literacy Challenges & Opportunities with Project-Based Learning Toolkit in K-12 Education (AAAI 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.