Research map / Vision-language and multimodal / Human and embodied modeling
ImageSTEAM: Teacher Professional Development for Integrating Visual Computing into Middle School Lessons
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.
Describes ImageSTEAM, a teacher professional development program co-creating middle-school lessons on computer vision, machine learning, and computational cameras, evaluated with pre/post teacher self-efficacy surveys.
Problem
Teachers lack training and interdisciplinary preparation to teach AI and visual computing in K-12 (middle school grades 6-8), creating a barrier to effective AI lessons.
Approach
A co-creation teacher PD model in which teachers build standards-aligned visual computing lessons, pilot them with middle school students, and iterate, supported by low-barrier web tools and custom Python/Colab notebooks.
Key results
- Teacher self-efficacy (T-STEM CT) rose from mean 18.58 to 29.5 pre to post (reported as 58.8% increase) in the 2022 cohort of n=12.
- Outcome expectancy rose from mean 20.33 to 22.42 (reported as 10.3% increase), a smaller gain than self-efficacy.
- 43 teachers and 189 students participated across summers 2021-2023 in Arizona and Georgia.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Participate in ImageSTEAM teacher professional development | Middle school teachers · 2022 summer cohort | self-efficacy score: 18.58 -> 29.5; 58.8% increase | Helped |
| Participate in ImageSTEAM teacher professional development | Middle school teachers · 2022 summer cohort | outcome expectancy score: 20.33 -> 22.42; 10.3% | Helped |
Surprising findings
- Math teachers were concerned that AI activities would take time from fractions and scale-factor instruction, and test pressure made math the hardest subject to integrate.
- Teachers and students found bias and ethics discussions less intimidating than technical AI/CV/ML topics.
Limitations
- Small sample (n=12) for the quantitative pre/post survey, so differences could not be statistically validated.
- Free generative tools like GauGAN were deprecated, limiting lesson sustainability.
- Lesson deployment lasted only one week, so student knowledge gains were hard to capture.
Experimental setup
Datasets: Hurricane Harvey post-flood satellite image dataset (10,000 images)
Metrics: T-STEM CT self-efficacy score T-STEM CT outcome expectancy score descriptive statistics (mean, SD, min, max)
Builds on: AI4K12 guidelines T-STEM CT survey Personal Construct Theory STEAM education
Closest papers by meaning
- Advancing Research on Equitable AI Education Through a Focus on Implementation: Insights from a Middle School Computer Vision Module Beta-Test (AAAI 2025 · similarity 0.88)
- An Effectiveness Study of Teacher-Led AI Literacy Curriculum in K-12 Classrooms (AAAI 2024 · similarity 0.86)
- Understanding K-12 Teachers’ Needs for AI Education: A Survey-Based Study (AAAI 2025 · similarity 0.86)
- “From Unseen Needs to Classroom Solutions”: Exploring AI Literacy Challenges & Opportunities with Project-Based Learning Toolkit in K-12 Education (AAAI 2025 · similarity 0.86)
- Empowering Educators in AI: Insights from Co-Designing an AI Microcredential with and for K-12 Educators (AAAI 2025 · similarity 0.85)
- Foundations of Autonomous Vehicles: A Curriculum Model for Developing Competencies in Artificial Intelligence and the Internet of Things for Grades 7–10 (AAAI 2024 · similarity 0.85)
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.