Diagnose Simpson’s Paradox in Retention
Use cohort analysis for mix-shift diagnosis.
A team needs to analyze mix-shift diagnosis, but the first draft uses a blended average and an unclear denominator. Build from decision to cohort start, cohort age, stable denominator, and next action. The common trap is changing the denominator or choosing a start event before users are eligible for the measured behavior. 1 Write the decision in one sentence. This prevents the chart from becoming decorative. 2 Choose the shared start event and assign every member to a cohort row. The start event creates the clock. 3 Compute cohort age and count the outcome at each age. Age columns make…
Sign up free — one personalized lesson every day, matched to your role and goals.
Already have an account? Sign in