Quiz: The Archaeology of Error

Test your understanding before moving on. Target: 70% or higher to proceed confidently.

Section 1: Multiple Choice (1 point each)

1. What is the primary distinction between individual cognitive bias and a systemic failure mode?

  • A) Cognitive biases are real; systemic failure modes are theoretical
  • B) Cognitive biases are properties of individual minds; systemic failure modes are emergent properties of institutions and systems
  • C) Cognitive biases affect everyone equally; systemic failure modes only affect experts
  • D) Cognitive biases are correctable; systemic failure modes are permanent
Answer **B)** Cognitive biases are properties of individual minds; systemic failure modes are emergent properties of institutions and systems. *Why B:* The chapter's central argument is that the failure modes it examines operate at the institutional level, not the individual level — they are features of how knowledge-producing systems work, not features of how individuals think. *Why not A:* Both are real phenomena. *Why not C:* Both can affect anyone, but systemic failure modes affect entire fields. *Why not D:* Both are correctable, though systemic failure modes require structural changes. *Reference:* Section 1.2

2. In the peptic ulcer case, why did the gastroenterology establishment resist Marshall and Warren's findings?

  • A) Because Marshall and Warren's evidence was genuinely weak
  • B) Because the establishment was engaged in a deliberate conspiracy to suppress the truth
  • C) Because the structural incentives of the field — career investment, textbook inertia, pharmaceutical economics — made resistance the locally rational response
  • D) Because bacteria cannot survive in stomach acid
Answer **C)** The structural incentives made resistance locally rational. *Why C:* The chapter emphasizes that resistance was not about stupidity or conspiracy but about structural forces that made skepticism the rational individual response, even though the system-level outcome was harmful. *Why not A:* The evidence was strong, including Marshall's self-experiment. *Why not B:* The chapter explicitly argues against conspiracy explanations. *Why not D:* While this was the initial belief, it was proven wrong by the evidence. *Reference:* Section 1.1, "What It Looked Like From Inside"

3. Which of the following is Stage 5 ("Resistance") in the lifecycle of a wrong idea?

  • A) A wrong idea is first proposed by a prestigious researcher
  • B) Counter-evidence is dismissed using institutional mechanisms like peer review and funding decisions
  • C) The field rewrites its history to make the correction seem inevitable
  • D) The idea becomes embedded in textbooks and training programs
Answer **B)** Counter-evidence is dismissed using institutional mechanisms. *Why B:* Stage 5 is when defenders of the consensus use institutional mechanisms — peer review, funding, hiring, conference selection — to suppress challengers. *Why not A:* That's Stage 1 (Introduction). *Why not C:* That's Stage 7 (Revision). *Why not D:* That's Stage 3 (Entrenchment). *Reference:* Section 1.3

4. The chapter argues that the 2008 financial crisis demonstrated:

  • A) A single failure mode (precision without accuracy) operating in isolation
  • B) Multiple failure modes operating simultaneously — authority cascades, incentive misalignment, survivorship bias, consensus enforcement, and false precision
  • C) That mathematical models are inherently unreliable
  • D) That economists were incompetent
Answer **B)** Multiple failure modes operating simultaneously. *Why B:* The crisis is presented as an example of the "full failure mode stack" — multiple structural forces interacting. *Why not A:* The chapter identifies at least five failure modes at work simultaneously. *Why not C:* The issue isn't mathematics itself but the assumptions underlying the models. *Why not D:* The chapter explicitly argues against individual blame. *Reference:* Sections 1.3, 1.5

5. According to the chapter, why were these six specific anchor examples chosen?

  • A) They are the most dramatic stories of failure
  • B) They span different fields, stakes, correction timelines, and methodologies — demonstrating universality
  • C) They are the only well-documented cases of knowledge failure
  • D) They all occurred in the 20th century
Answer **B)** They span different domains, demonstrating that structural failure modes are universal. *Why B:* Section 1.5 explicitly states the selection criteria: different fields, different stakes, different timelines, all resisting the "stupid people" explanation. *Why not A:* Drama wasn't the criterion; structural representativeness was. *Why not C:* There are many other cases; these were selected for their diversity. *Why not D:* Some (bloodletting, dietary fat) span much longer periods. *Reference:* Section 1.5

6. What is the "revision myth" (Stage 7)?

  • A) The tendency to revise wrong conclusions quickly
  • B) The process by which fields rewrite their history to make corrections seem inevitable, making the next error harder to catch
  • C) The myth that scientific revisions are always improvements
  • D) The idea that history always repeats itself
Answer **B)** Fields rewrite history to make corrections seem inevitable. *Why B:* The revision myth creates the illusion that the system is naturally self-correcting, which makes current errors harder to recognize. *Reference:* Section 1.3

7. The chapter argues the book is NOT:

  • A) An argument against expertise
  • B) An argument for structural reform of knowledge-producing institutions
  • C) A systematic examination of how wrong ideas persist
  • D) A framework for diagnosing failure modes
Answer **A)** An argument against expertise. *Why A:* Section 1.8 explicitly lists what the book is NOT, including "an argument against expertise." *Reference:* Section 1.8

8. The threshold concept in this chapter is:

  • A) The seven-stage lifecycle
  • B) The structural nature of epistemic failure — being wrong is about systems, not stupidity
  • C) The six anchor examples
  • D) The Epistemic Audit methodology
Answer **B)** The structural nature of epistemic failure. *Why B:* The threshold concept box in section 1.2 identifies this transformation: from "we need better people" to "we need better systems." *Reference:* Section 1.2

Section 2: True/False with Justification (1 point each)

9. "The existence of failure modes in expert consensus means we should generally distrust experts."

Answer **False.** The chapter explicitly argues that this is the most dangerous misreading of the book. Failure modes are real but the same structures that produce wrong consensus also produce correct consensus. The goal is better diagnostic tools, not blanket distrust. (Section 1.8)

10. "The seven-stage lifecycle of a wrong idea always proceeds in strict linear order."

Answer **False.** The chapter notes that "not every wrong idea passes through every stage, and the timeline varies enormously." The lifecycle is a diagnostic pattern, not a rigid sequence. (Section 1.3)

11. "A field that has corrected a past error is unlikely to make similar errors in the future."

Answer **False (mostly).** The revision myth (Stage 7) actually makes future errors *harder* to catch by creating the illusion that the system is self-correcting. Fields that sanitize the history of their past corrections are, if anything, more vulnerable to the next error. (Section 1.3)

12. "The structural account of failure modes means individuals bear no responsibility for errors."

Answer **False.** The chapter argues the problem is "primarily structural" but "not exclusively" — individual biases are real and matter. The structural account doesn't eliminate individual responsibility; it explains why individual good intentions are insufficient to prevent systemic failure. (Section 1.2)

Section 3: Short Answer (2 points each)

13. Explain the concept of "locally rational resistance." Why is it important for understanding knowledge failure?

Sample Answer Locally rational resistance means that each individual's decision to resist new evidence was reasonable given their information, incentives, and career position. A senior gastroenterologist dismissing the bacterial ulcer hypothesis was acting rationally from their perspective: the existing model worked adequately, the challenger's credentials were modest, and acceptance would invalidate decades of their own work. The concept is important because it shows that systemic failure doesn't require bad actors — it can emerge from the aggregate of individually reasonable decisions within a poorly designed system. *Rubric — full credit requires:* - Definition of "locally rational" in context - Example demonstrating the concept - Explanation of why this matters for understanding systemic vs. individual failure

14. The chapter lists three principles for ethical use of the book's diagnostic tools: consistency, proportionality, and humility. Briefly explain each.

Sample Answer **Consistency:** Apply the same diagnostic standards to claims you want to be true and claims you want to be false. **Proportionality:** The existence of failure modes doesn't justify ignoring a field entirely; the question is whether the current system is the best available source of knowledge. **Humility:** Your own analysis (including your Epistemic Audit) is not immune to the failure modes it diagnoses — build in checks. *Rubric — full credit requires:* - Accurate description of all three principles - Some indication of practical application

Section 4: Applied Scenario (3 points)

15. A pharmaceutical company has published a study showing that their new drug reduces symptoms of a chronic condition by 40%. The study was funded by the company, conducted at a university with strong ties to the company, published in a high-impact journal, and has not been independently replicated. A competitor's drug, which has been available for 20 years and has extensive replication data, shows a 25% reduction.

Using the concepts from Chapter 1, analyze this scenario. Which failure modes might be relevant? What additional information would you need to evaluate the claim? What would you recommend?

Sample Answer Several failure modes are potentially relevant: **incentive misalignment** (the study was funded by the company that profits from positive results), **precision without accuracy** (a specific 40% figure from a single study may imply more certainty than warranted), **authority cascade** (publication in a high-impact journal and university association lend credibility), and **survivorship bias** (we don't know about unfavorable results that weren't published). Additional information needed: Has the study been independently replicated? What is the effect size when adjusted for industry funding bias? Were there pre-registered outcomes? What is the safety profile? How large and diverse was the sample? Recommendation: The competitor's drug with 20 years of replication data is currently a more reliable bet despite the smaller effect size. The new drug's 40% figure should be treated as provisional until independent replication confirms it. This doesn't mean the new drug is wrong — it means the evidence is insufficient to displace an established, well-replicated alternative. *Rubric:* | Criterion | 0 pts | 1 pt | 2 pts | 3 pts | |-----------|-------|------|-------|-------| | Failure mode identification | None identified | One identified | Two identified correctly | Three+ identified with explanation | | Evidence evaluation | No analysis | Surface analysis | Identifies key uncertainties | Systematic analysis with specific questions | | Recommendation quality | None or inappropriate | Reasonable but vague | Specific and well-reasoned | Nuanced, acknowledges uncertainty |

Scoring & Next Steps

Score Assessment Recommended Action
< 50% Needs review Re-read sections 1.1–1.3 and the chapter summary, redo Part A exercises
50–70% Partial Review weak areas, focus on the lifecycle and failure mode taxonomy
70–85% Solid Ready to proceed; revisit any missed topics
> 85% Strong Proceed to Chapter 2; consider the Deep Dive extensions