Chapter 6 — Quiz

Twenty-two questions.

Multiple choice

1. The first stage of the hype pipeline is: a) a testimonial b) a real preclinical result c) a vendor appearing d) media coverage

2. What is lost between stage 1 and stage 2? a) the funding source b) the model, species, dose, and endpoint c) the peer review d) the author names

3. The single most epistemically honest text in the pipeline is typically: a) the abstract b) the press release c) the authors' own limitations paragraph d) the forum summary

4. Regression to the mean predicts that people who start a treatment at their worst point will: a) get worse b) stay the same c) improve on average regardless of the treatment d) show no pattern

5. Survivorship bias in testimonials means: a) only survivors of side effects post b) the visible reports are filtered for positive outcomes before you see them c) old posts outrank new ones d) vendors delete negative reviews

6. "Research chemical — not for human consumption" primarily accomplishes: a) protecting purity b) avoiding drug regulation and liability while permitting a human-use market c) meeting shipping requirements d) reducing cost

7. Epistemic laundering describes: a) falsifying data b) assembling a claim from components so that no participant states it c) removing a paper from a journal d) paying for peer review

8. Media coverage of peptides typically arrives in: a) one accurate register b) two registers, miracle and menace, both misleading c) purely negative coverage d) purely positive coverage

9. The chapter argues the "menace" register does specific damage because: a) it is legally risky b) it teaches readers already using a compound that mainstream coverage does not understand them, discrediting the accurate parts too c) it reduces sales d) it is always factually wrong

10. In the incentive diagram, who benefits from "the evidence is currently insufficient"? a) researchers b) vendors c) media d) nobody

11. Funding source, per Chapter 5's rule, changes: a) the truth value of evidence b) the weight of evidence c) nothing d) the legality of the claim

12. BPC-157's distortion enters primarily at the step of: a) fabricating data b) extrapolating from rodents to humans as if volume accumulated into human evidence c) misreporting the dose d) hiding side effects

13. NAD+ precursors differ from BPC-157 in that: a) they have no supporting science b) human trials exist and are unimpressive, rather than absent c) they are approved drugs d) they are peptides

14. For GHK-Cu, the decisive unaddressed question is: a) whether the peptide exists b) whether it affects matrix production in the laboratory c) whether it penetrates intact skin in meaningful quantity d) whether copper is toxic

15. "Do your own research" fails in this environment because: a) people cannot read b) the accessible information has been pre-sorted by ranking and commercial optimization c) primary sources are paywalled d) research is always inconclusive

16. Mechanistic understanding affects susceptibility by: a) decreasing it reliably b) having no effect c) increasing it, because a vivid mechanism feels like evidence d) increasing it only in non-experts

17. The fastest disqualifier in the sixty-second protocol is: a) checking the funding b) asking what would falsify the claim c) reading the abstract d) checking the dose

Short answer

18. Name the five mechanisms that make a testimonial uninformative, and identify which is structural rather than psychological.

19. Explain how a vendor page can be legally accurate and communicatively misleading at the same time.

20. Explain why the honest version of a trial result cannot be a headline, and why this is structural rather than a failure of individual journalists.

Applying the rating discipline

21. Write the four-line rating for NAD+ precursors for human healthspan extension, then explain why it is ⚠️ rather than ❌ despite the strong version of the claim being unsupported.

22. §6.7 states that this book has its own biases, naming a preference for rigor that can become excessive conservatism. Which of Chapter 5's six rating rules is that a threat to, and what procedural check would detect it?

Answer key **1.** b. **2.** b. **3.** c. **4.** c. **5.** b. **6.** b. **7.** b. **8.** b. **9.** b. **10.** d. **11.** b. **12.** b. **13.** b. **14.** c. **15.** b. **16.** c. **17.** b. **18.** Regression to the mean; natural history; expectation; co-intervention; **survivorship**. The first four are psychological or biological properties of the individual case. **Survivorship is structural** — it operates on the *record you see* rather than on the person reporting, filtering the visible sample for positive outcomes before you arrive. It is the same problem as publication bias, operating on testimonials, and it is the one you cannot correct for by being a careful reader of any individual post. **19.** The disclaimer "we make no medical claims" may be entirely accurate as a legal statement — the vendor really is not asserting that the compound treats anything. But the page is constructed from elements that collectively produce a therapeutic belief: links to healing studies, a purity certificate implying the important question is settled, a dosage calculator, injection supplies as related products, and customer reviews describing outcomes. **Each element is individually defensible; the assembly is the claim.** The test is what the page is *for* — a reagent supplier selling to laboratories needs none of that infrastructure. **20.** The honest statement of a result requires the relative figure, the absolute figure, the population, the duration, and the comparator — roughly three times the length of "cuts events by 20%," and it resolves to something qualified rather than dramatic. This is structural rather than a failure of individuals: formats that reward engagement systematically favor the shorter, more definite version, and the effect operates identically on drugs the writer likes and drugs the writer distrusts. Fixing it requires the reader to want the longer version, which at scale they do not. **21.** **Claim:** Supplementation with NAD+ precursors extends healthspan or produces meaningful functional benefit in humans. **Rating:** ⚠️ Promising but preliminary, trending toward ❌ for the strong version. **Why:** the biochemistry is real and the precursors demonstrably raise NAD+ in humans — but that is a surrogate; human trials have been mostly small and short with modest and inconsistent functional effects. **What would change it:** an adequately powered, long-duration randomized trial with a pre-specified functional or clinical endpoint, not a biomarker. **Why ⚠️ and not ❌:** because human trials *exist*. ❌ is reserved for claims supported by animal or in-vitro data with no completed human trials, or contradicted by human trials. Here there is real human data that is genuinely inconclusive — which is a different state from "untested," and the rating system distinguishes them deliberately. Note that this makes NAD+ precursors' position *epistemically stronger* than BPC-157's while being *commercially weaker*, since "tested and unimpressive" is harder to sell than "the studies are amazing." **22.** It threatens **rule 4 — never downgrade a rating with distaste.** A preference for rigor that shades into excessive conservatism produces systematically harsher ratings for compounds with low-quality evidence bases *beyond* what the evidence justifies — and it feels like scrupulousness from inside, which is exactly why it is hard to detect. **The procedural check:** apply the six trial-quality questions and the four-tier criteria mechanically, and see whether the *answers* differ between compounds you respect and compounds you don't, rather than whether the conclusions do. If two compounds have the same answers on all six and different ratings, something other than the evidence is doing the work. Chapter 37's "find your drift" exercise applies this to the reader's own dossier, and it applies to this book too.