TRESTLEIndependent minds, one table.
Learn to build the quantitative fluency — confidence intervals, sample sizes, defensible comparisons — that makes a UX researcher’s findings hard to dismiss, in plain language, at your own pace.
Get access — new learners just enter their email to start, no password needed.
How it works
From someone who’s been exactly where you are.
Trestle exists because the knowledge that makes a finding hold up shouldn’t be locked behind a $2,000 course or a credential most people don’t have. It’s built by someone who had to fight for that knowledge the hard way — and decided no one else should have to.
A real clip from Phase 1 — the Wilson score interval, explained.
Method: Wilson score interval (Wilson, 1927). UX application: Sauro & Lewis (2016).
What you’ll learn
Phase 1 — Can You Trust This Number? (Free)
This is the method you reach for whenever someone asks:
- “Can we really say 8 out of 10 users succeeded?”
- “Why can’t 5 users give us a precise number?”
- “Why do we need 70+ people for this study?”
- “How many participants do I need to actually catch a real improvement, not just noise?”
- “Before I field this survey — how much precision am I actually buying with this sample size?”
Covers confidence intervals, sample-size reasoning, and power analysis — so you can answer these with real statistics, not a shrug.
Phase 2 — Which Design Actually Wins? ($50)
Once you can trust a single number, the next question is comparing two. This is the method for:
- Comparing time-on-task, error rate, or satisfaction across two or more prototype directions
- “Design B tested faster — but is that a real difference, or just noise?”
- Telling a PM “yes, B is faster, and here’s how much faster — not just that it’s significant”
Covers t-tests, ANOVA, and effect size.
Phase 1 is free. More phases are rolling out through 2026.
Want to talk it through with other learners first?
Join the community (optional)Free · Discord · Open to all Trestle learners