Exercises: The Epistemic Health Checklist

Part A: Comprehension and Application

A.1. For each of the 10 dimensions, identify the primary failure mode from Parts I-III that it detects. (Example: Dimension 1, Dissent Tolerance, detects consensus enforcement from Chapter 14 and the outsider problem from Chapter 18.)

A.2. Explain the difference between the Red Flag Scorecard (Chapter 31) and the Epistemic Health Checklist. When would you use each tool? Can they be used together?

A.3. The chapter argues that "the profile pattern matters as much as the average." Explain this claim with an example. Why might a field with an average score of 5 and an uneven profile be more dangerous than a field with an average score of 4 and a uniform profile?

A.4. Score criminal justice (Chapter 27) on all 10 dimensions using the indicators in this chapter. Compare your scores with the nutrition science worked example. Where do the two fields' profiles differ, and what structural features explain the differences?

A.5. Score the military (Chapter 28) on all 10 dimensions. The military has strong learning infrastructure (war colleges, after-action reviews) but still repeats errors. Which dimensions capture this paradox?

Part B: Analysis

B.1. The worked examples score medicine at 6.0 (mixed), nutrition at 3.1 (poor), and software engineering at 6.1 (mixed). Apply the checklist to your own field and compare your field's profile to these three. What does the comparison reveal about your field's specific vulnerabilities?

B.2. Dimension 3 (Incentive Alignment) scored 2/10 for pre-2008 finance. Apply this dimension to a current industry or field of your choice. What specific incentives reward error-producing behavior? What incentives reward truth-seeking? Which dominates?

B.3. The chapter excludes "crisis probability" as an explicit dimension. Design an 11th dimension that captures crisis-related correction capacity. Define the scoring indicators (green/yellow/red) and calibrate it against at least three fields from Part IV.

B.4. Two organizations in the same field score the same average on the checklist but have different profiles — one is uniformly moderate (all 5s) and the other is highly uneven (three 8s and seven 3s). Which organization is in greater danger? Design a scenario that illustrates why.

Part C: Synthesis and Evaluation

C.1. The Epistemic Health Checklist is designed to assess fields and organizations. Could it be adapted to assess yourself — your own epistemic health as an individual thinker? Design a personal version with 10 dimensions. Which dimensions would change from the field version, and which would stay the same?

C.2. A university provost asks you to assess the epistemic health of their institution — not a field, but a specific organization. How would you adapt the checklist? Which dimensions translate directly, and which need modification for organizational rather than field-level assessment?

C.3. Evaluate the Epistemic Health Checklist itself using the Red Flag Scorecard from Chapter 31. Apply at least 5 of the 15 red flag questions to the checklist's claims. Does the tool pass its own test?

Part D: Mixed Practice (Interleaved)

D.1. A pharmaceutical company claims that its internal research culture is epistemically healthy — it funds replications, tolerates dissent, and corrects errors quickly. Apply both the Epistemic Health Checklist AND the Red Flag Scorecard to evaluate this claim. What do the two tools reveal when used together?

D.2. You are advising a new research institute that wants to avoid the failure modes documented in this book. Using the Epistemic Health Checklist, design the institute's structures — hiring, funding, publication, dissent mechanisms — to score at least 7/10 on every dimension. What trade-offs are required?

D.3. Compare the Epistemic Health Checklist to the Correction Speed Model (Chapter 22). Both assess a field's correction capacity, but they use different variables. Map the 10 Checklist dimensions onto the 8 Correction Speed Model variables. Where do they overlap? Where do they diverge? Which tool is more useful for which purpose?