Blog

When Your Ruler Is Wrong: How Rescoring Improves Measurement in LEVANTE

How we validate translations

Starting with traditional review and backtranslation The Promise of Embeddings Using Gemini to compare meaning, not just strings Why we use multiple tools for validation Where we are heading next

LEVANTE Data Release and Hackathon — Join us this summer!

Expanding data access What’s available: Supporting innovation through data education Week 1: Open virtual sessions Week 2: Project-based hackathon (application required) Get involved

From Averages to Individuals: The Next Frontier of Evidence-Based Education

The Problem with “On Average” Where LEVANTE can help

Introducing the LEVANTE core tasks: open, cross-cultural measures for measuring learning and development (ages 5–12)

Why is measurement such a bottleneck in developmental science? What LEVANTE is trying to solve Where we’re going next Join us!

Bringing child development research online at scale

How does the dashboard work? Privacy and ethics

Speaking Every Child’s Language:  Ensuring authentic translation of LEVANTE tasks across countries

Our translation process starts with automated “AI” tools AI tools are only a starting point, not an end point Keeping task difficulty consistent Working towards cultural relevance Grammatical differences between languages More to do

Global Collaboration to Capture Variability in Child Development

The Challenge Our goal This blog References