UpGrade¶
Product or Programme · All products and programmes
Evidence · 3 claims (3 for) · 3 studies (2 causal, 1 qualitative),q2· 0 of 3 report an effect size · 3 claims rest on one study
Description¶
UpGrade is a free, open-source platform developed for conducting field trials in educational technology software, including experiments integrated with Carnegie Learning's MATHia adaptive mathematics tutoring system.
Components¶
- MATHia/UpGrade: open-source field-trial platform integrated with Carnegie Learning's adaptive math tutoring system: UpGrade is "a free and open source platform for conducting field trials in EdTech software" integrated into MATHia, "an adaptive, intelligent tutoring system for learning math" used by over 600,000 students in over 3,000 schools each year. Researchers run content, feature, and sequence-based experiments in real classrooms, with weighted randomization or factorial designs, by-experiment outcome metrics, dashboard snapshots, and csv data exports. The article illustrates rewriting word problems for clarity and relatable scenarios such as a Taylor Swift security-guard problem. (Language and Mathematics Learning: A Comparative Study of Digital Learning Platforms)
- UpGrade open-source A/B testing platform for classroom-embedded field experiments: UpGrade, developed by Carnegie Learning in 2020, is an open-source platform for conducting field trials in educational software. It acts as a middle layer between a client application and the experimenter, managing "scheduling, randomization, and parameterization of all experiments (and feature flags)" via three key API calls and decision points. It supports individual and group random assignment and lets experiments enroll students as they encounter target content in adaptive curricula. To date it has been used in over 73 field experiments impacting hundreds of thousands of students. (Classroom-Based Experimentation in a Digital Learning Platform)
Claims¶
- Localizing word-problem character names improved MATHia performance only for specific subgroup-majority schools, with no improvement in sense of belonging overall [+W]
- Rewriting MATHia word problems for struggling readers, by human experts or LLMs, sped completion by 30% and improved mastery rate [+W]
- Carnegie Learning attributes positive practitioner feedback to connecting research to participants' needs, lowering effort barriers, and compensating participation [+W]
Related Products and Programmes¶
-
Key Sources¶
- Wei, X., Wortman, A., Cheng, L., Heffernan, N., Heffernan, C., Murphy, A., Zepeda, C., Motz, B., Jankowski, H., & Roschelle, J. (2024, March). Language and mathematics learning: A comparative study of digital learning platforms. Digital Promise. https://doi.org/10.51388/20.500.12265/206
- Murphy, A. (2026, July). Classroom-Based experimentation in a digital learning platform. Digital Promise. https://doi.org/10.51388/20.500.12265/306
Page metadata
| Field | Value |
|---|---|
type |
product |
id |
upgrade |
title |
UpGrade |
description |
UpGrade is a free, open-source platform developed for conducting field trials in educational technology software, including experiments integrated with Carnegie Learning's MATHia adaptive mathematics tutoring system. |
status |
review |
moved_from |
element/mathia-upgrade-field-trial-platform, element/upgrade-ab-testing-platform |
product_kind |
software |
generated |
by process:sweep-kinds on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
sources |
2 — wei-x-wortman-a-cheng, murphy-a-2026-july-classroom-based |
Frontmatter is the page's machine-readable record — see the Schema & Guide.