Field Testing of Teacher-Driven Innovations in EdLight

EdLight, PBC

⚡ AT A GLANCE

RESOURCE TYPE *

Platform/Experiment

POPULATION FOCUS *

Students Teachers

DATA AVAILABLE *

Quantitative Standards-Aligned

LAST UPDATED (DATE) *

2026 January 26

SUBJECT AREA(S) *

MATH

EDUCATIONAL LEVELS COVERED *

6-8

PLATFORM DESCRIPTION (BRIEF) *

The EdLight platform supports practitioner-driven research and research-practice partnerships (RPP). Practitioners and researchers can co-design tagging schemas, upload new tasks or prompts for annotation, and conduct experiments within the classroom context. This option is designed for active field testing, allowing for the evaluation of new instructional interventions and real-time student work analysis with or without formal research partners.

DOWNLOADS AND LINKS

📄

Data dictionary

🎬

Platform Demo

💡

SAMPLE RESEARCH QUESTION

What is the impact of making model-generated feedback visible to teachers on the quality and speed of instructional responses?

1. Capabilities & Content

Resource Type *

Platform/Experiment

Supports Experimental Research? *

Yes – full experimental control

Instrumented Capabilities *

Experimental Control

Random assignment to groups Control timing of interventions

Content Manipulation

Inject new learning materials

Intervention Deployment

Chatbot or conversational agent

Data Collection Extensions

Custom event logging External survey integration

Additional Notes on Capabilities / Customization

Researchers and practitioners can upload tasks or prompts for annotation and customize tagging schemas to align with their instructional or research focus. Researchers and practitioners may request custom tagging workflows for comparative content analyses.

The platform itself does not deploy interventions but can be used to evaluate the impact of instructional interventions already implemented in classrooms, particularly when aligned with curriculum-embedded tasks.

Researchers and practitioners can request extensions such as:

• Customized tagging of new student work uploads (e.g., curriculum-specific error patterns)

• Alignment to new standards or instructional frameworks

• Addition of researcher-generated annotations (via structured protocols)

• Integration of external datasets via joinable identifiers (e.g., school IDs, curriculum types)

Limitations include:

• No direct video/audio capture or behavioral observations

• Annotation schema is optimized for math and may not generalize to other subjects

Customizations require approval through EdLight’s secure research onboarding process.

Data Fields (Demographic Options)

None

Linkable Identifiers Present? *

Yes

If 'Yes' or 'Partial', List Relationship Details

Datasets described can be joined by student ID.

Data Type / Format

Clickstream Data Assessment Responses Standards-Aligned

2. Research Potential

Example Research Questions

What is the impact of making model-generated feedback visible to teachers on the quality and speed of instructional responses?

How does the visual expression of learning develop understanding of math concepts in new experimental groups?

How do student outcomes change when teachers use customized AI-tagging schemas for specific curriculum-embedded tasks?

Priority Research Areas *

School/System Improvement Assessment & Measurement Teacher Support & Orchestration Affect & Motivation Equity, Access & Inclusion

Construct(s) of Interest

Student Mathematical Thinking, Misconception Detection, Instructional Responsiveness, Equity of Access to Feedback, AI Interpretability and Trust, Data-Driven Teacher Decision-Making

3. Related Research Examples

Fostering Teacher & Student Growth

Internal Report

Valley View School District (VVSD), just outside of Chicago, Illinois, has distinguished itself with a strong commitment to equity and academic excellence, undergirded by the belief that students should be centered as the knowledge-holders within the classroom. EdLight has worked hand-in-hand with VVSD to ensure the district is able to put their values into practice.

Catalyzing Student Achievement

Internal Report

Catalyst Public School is a K-12 school, free of tuition and admission requirements, committed to fostering a nurturing and innovative learning environment where every student can thrive. Their uncompromising focus on holistic education ensures that students are not only academically proficient but also emotionally and socially equipped to face the challenges of tomorrow. EdLight, with its easy-to-use, student-centered platform and AI-enabled insights on student work, was just the right partner bring Catalyst’s values to life.

Using Student Work Data to Improve Teaching and Learning

Internal Report

This report explores the potential ways that student classwork in middle grades mathematics can serve as a more timely, nuanced, and detailed source of evidence on student progress when collected at scale. We draw on a unique data set, collected and processed by EdLight, that pulls together a comprehensive panel of student class assignments and uniformly tags these assignments through the combination of expert and AI review. Our report explores emerging patterns from this analysis of student classwork and highlights connections between EdLight’s student classwork data and traditional measures to gain deeper insights into how overall student classwork ratings, the strategies students use, and the errors students make correlate with student performance.

4. Funding Opportunities

Open Grants

Open

AIMS EduData Spring 2026

Award Range: $10,000 - $400,000 Due: 2026-03-02

View Details

5. Access & Collaboration

Access Process

Expression of Interest IRB Required Consultation

Data Access Process / Timeline

1. Partnership discovery call to define experimental goals.

2. Co-design of tagging schemas and task selection.

3. Classroom/District implementation setup.

4. Real-time data collection and iterative analysis (4-8 week onboarding).

External researchers are supported through structured onboarding, Q&A sessions, and data office hours.

All image data is de-identified prior to access.

Use cases with minors include IRB and district approvals where required.

Role-based access ensures only authorized researchers can view or export subsets.