Test design
Experiment Design Clinic
Small groups rotate through facilitated critiques of real experiment docs. Expect redlines on scope, population definition, and how success is communicated post-readout.
- Cadence
- 4 live sessions
- Format
- Cohort
- Funnel framing
- Consideration
- Level
- Intermediate
- Tuition indication
- 1,120,000 KRW
Practice modules
- Hypothesis one-pager template
- Guardrail metric checklist
- Sample size sanity pass with transparent assumptions
- Storytelling kit for neutral readouts
- Pair exercise on conflicting north-star metrics
- Office hour for statistics questions
Measured takeaways
- Submit two experiment plans ready for review
- Agree on language for inconclusive results
- Reduce approval churn by front-loading risks
Mentor on point
Amelia Cho
Analyst mentor with background in regulated consumer apps.
Field notes FAQ
Is Bayesian testing covered?
We reference it where teams already use those tools, but frequentist framing is the default.
What if leadership wants faster reads?
We practice trade-off conversations and staged releases instead of shortening rigor silently.
Are templates shareable?
Yes for internal use. Redistribution outside your organization requires written permission.
Voice from recent cohort desks
Stopped us from stuffing twelve metrics into every doc. Guardrail sheet is now default.
Tone is pragmatic; would like a deeper multi-armed bandit appendix next time.
