Discussion Guide — Chapter 23

Six prompts. Each is written to have no comfortable answer; if the room converges quickly, the question was misheard and you should say so.

1. "Suppose you learn tomorrow that forty well-conducted trials of Semax exist in Russian, and that none has been translated. Does your rating change?"

What to listen for: the correct answer is no — the rating describes what is evaluable, and learning that inaccessible evidence exists does not make it assessable. Students who say yes have converted Ledger B into Ledger A, which is exactly the silent migration §23.3 warns about. But listen also for the sophisticated version: the access flag absolutely should change, because now you know the size of the gap. A student who separates those two responses has understood the chapter. Push anyone who says "well, forty is a lot" to say what forty of what would be.

2. "Rule 4 says never downgrade with distaste. Is unfamiliarity really the same thing as distaste?"

What to listen for: this is a genuine objection and deserves a real hearing. Distaste is an affective response; unfamiliarity is an epistemic state. The chapter's claim is that they behave identically in a rating — both cause a downgrade for a reason that is not a property of the evidence. Strong students will notice the asymmetry that makes unfamiliarity more dangerous: it feels neutral. Watch for the student who tries to rescue unfamiliarity by arguing it correlates with quality; ask them for the evidence, and then ask whether they can assess that evidence.

3. "The chapter says a Western regulator's non-approval is usually a commercial decision. Doesn't that prove too much? Couldn't you use that argument to excuse any unapproved compound?"

What to listen for: yes, and that is the correct worry. The argument establishes only that non-approval is weak evidence about efficacy; it does not establish that any particular compound works. The good discussion distinguishes this claim is not evidence against from this claim is evidence for. Students who cannot make that distinction will make it later in their lives about something that matters more. Push toward the asymmetric case: when is non-approval informative? (When a well-resourced sponsor had incentive and opportunity, filed, and was refused.)

4. "Cerebrolysin has accessible trials that disagree, and Semax has inaccessible trials. Which situation would you rather be in as a reader — and which as a patient?"

What to listen for: the split between the two roles is the payload. As a reader, disagreement is enormously more informative than silence: you can characterize the effect's upper bound. As a patient or clinician, neither is satisfying, and the practical considerations (approval status, monitoring, product identity) may dominate the evidentiary ones entirely. Students who answer the same way for both roles have probably not thought about the second one. Do not let the discussion drift into recommendations — redirect to what each situation lets you say.

5. "If an effect only appears when you know you took the compound, is it a real effect?"

What to listen for: expect a fast "no" and push back. Something real happened: the person did perform better. The chapter's position is that it is real in the person and is not a property of the molecule — and the molecule is what is being sold, at a price, with risk. The best discussions get to the uncomfortable question of whether that distinction matters to someone who just wants to concentrate, and the honest answer is that it matters for the purchase decision and the risk assessment even if it does not matter for the experience. Do not resolve this cleanly; it should itch.

6. "The chapter claims its standard is not a Western standard. Attack that claim."

What to listen for: the best attacks are infrastructural rather than philosophical — pre-registration presupposes registries, adequate powering presupposes funding, publication-regardless-of-outcome presupposes venues willing to publish nulls. These are real and they complicate applying the standard. The distinction the class should reach: a standard can be universally valid and unevenly affordable, and those require different responses. Anyone who concludes "therefore the standard is arbitrary" has overshot; ask them which of the eight items they would drop, and what would then be believable.