Lessons
1When the average tells the wrong comparison35 min read
Calculate subgroup and aggregate rates to expose a composition effect.
- →Diagnose an aggregate reversal from subgroup counts
- →Separate subgroup composition from within-group performance
2Critique a result without substituting your own story35 min read
Build an evidence audit with claims, support, and missing tests.
- →Critique a research claim with a structured evidence audit
3Propose a study that can fail informatively35 min read
Define a small research proposal with alternatives and a stop rule.
- →Propose a bounded study with discriminating outcomes
4Defend a research talk under a changed assumption35 min read
Revise a conclusion when an interviewer changes the evidence conditions.
- →Defend a research contribution and revise a claim under challenge
- →Explain uncertainty without losing the decision
Skills in this course
- 01Diagnose an aggregate reversal from subgroup countsDiagnose an aggregate reversal from subgroup counts.
- 02Critique a research claim with a structured evidence auditCritique a research claim with a structured evidence audit.
- 03Propose a bounded study with discriminating outcomesPropose a bounded study with discriminating outcomes.
- 04Defend a research contribution and revise a claim under challengeDefend a research contribution and revise a claim under challenge.
- 05Separate subgroup composition from within-group performanceSeparate subgroup composition from within-group performance.
- 06Explain uncertainty without losing the decisionExplain uncertainty without losing the decision.