Evaluation · Shape
Evaluation · ShapeInteractive Evaluation
A two-layered measurement practice: A/B tests for short-term causal attribution, longitudinal cohorts for behavioural drift over weeks and months.
Why it matters
A sustained view of whether the AI is really helping, catching regressions that are invisible at the statistical-significance layer.
The canvas
The worksheet you fill in, as a template. Use it with your team here, or fill it in online in the workspace.
Project Team Date
User-outcome KPIs: did the user achieve their job?
A/B: which one variable will you test first?
Longitudinal: which cohort, re-run every 6 weeks?
What would drift look like in your data?
Explanation audit: when, and against which cognitive-load heuristics?
Who reviews this, and which unintended consequences have you mapped?
app.jcerejo.com · The Anticipatory Design Framework