Discussion Guide — Chapter 21

Six prompts, roughly in chapter order. Each includes what to listen for and the failure mode to watch for.

Prompt 1 — "Oxytocin is on the WHO Essential Medicines List and is also the subject of the most misleading nickname in popular neuroscience. Is that a contradiction? If not, what exactly is it?"

Open with this. It sets the frame for everything else and immunizes the room against the credulous-to- cynical flip.

Listen for: rating rule 6 (one molecule, many ratings) and rule 1 (ratings attach to claims with a population and endpoint). Strong answers notice that the two statements concern different routes, different populations, and different endpoints, and that a single verdict on "oxytocin" would destroy exactly the information a reader needs.

Watch for: students who resolve the tension by deciding one of the two statements must be softer than it looks. Both are stated at full strength and both hold.

Prompt 2 — "The prairie vole finding is that receptor placement, not the peptide, drives the species difference. Does that finding support or undermine the leap to human social behavior?"

Listen for: the recognition that it undermines it — the whole content of the result is that closely related species behave oppositely because their receptor maps differ, so it is evidence against casual cross-species inference, and it was cited in support of exactly that inference. This is the sharpest single point in the chapter and about a third of a class gets there unaided.

Watch for: students who conclude the vole work was therefore flawed. Redirect hard. The work was good; the citation of it was bad; these are different people making different errors at different times.

Prompt 3 — "A widely publicized literature reported that oxytocin increases trust. Larger studies mostly did not reproduce it. Reconstruct how that happened without anyone behaving dishonestly."

Listen for: small samples, noisy behavioral outcomes, the winner's curse, researcher degrees of freedom, publication incentives. The best answers explain that each individual decision was defensible and the aggregate was not — which is the general shape of most systemic scientific error.

Watch for: two failure modes in opposite directions. Some students want a villain and will invent one. Others will conclude the whole field is worthless, which the chapter explicitly refuses. Press the second group: what did psychology actually do in response? (Preregistration, registered reports, multi-lab collaboration, larger samples.) The correction is part of the story.

Prompt 4 — "Suppose we established tomorrow that intranasal oxytocin produces negligible central receptor occupancy. Which findings in this chapter would need re-explanation, and how?"

The most productive prompt in the set. Give it real time.

Listen for: students working out that this would compromise the positive findings and the negative ones, and that the context-dependence work — which the chapter presents as the more careful literature — uses the same route and would be equally affected. The realization that a methodological failure can undermine a correction as thoroughly as it undermines an error is the sophisticated destination here.

Watch for: the tense shift into "it has been shown not to reach the brain." The prompt is counterfactual. Hold them to it.

If the room is strong, run the inverse: suppose a PET study established substantial occupancy. Which ratings move? (Answer: the autism ❌ becomes interpretable as a genuine biological null rather than an ambiguous one, which arguably strengthens it. Most students expect the opposite, and working out why they were wrong is worth the detour.)

Prompt 5 — "The chapter says the ❌ for autism is 'evidence present and negative, not evidence absent,' and that this is a stronger scientific position and a worse commercial one. Unpack the second half."

Listen for: the market implication — that a compound which has never been tested is easier to sell than one that has been tested and failed, so the incentive structure rewards not running the trial. Strong answers connect this to Part III and note that absence of negative evidence routinely gets marketed as absence of problems.

Watch for: students treating this as a cynical aside. It is a structural claim about incentives and it predicts things. Ask them what it predicts about which compounds get trials.

Prompt 6 — "'A hormone named for one of its effects will be reasoned about as though that effect were its purpose.' Pick a nickname the chapter does not discuss and take it apart."

Close with this. It converts the chapter into a portable tool.

Listen for: specificity about what the compression loses. "Cortisol does more than stress" is a weak answer; "cortisol is essential to waking, to blood glucose regulation, and to suppressing inflammation, and the nickname makes people treat all three as side effects of a stress function" is a strong one. Good candidates: cortisol, dopamine, serotonin, leptin, testosterone, melatonin, adrenaline, and — the one that makes the chapter's ghrelin pairing land — insulin as "the fat-storage hormone."

Watch for: the answer that stops at "nicknames are oversimplifications." True and useless. Push for the mechanism of the error: the nickname does not merely omit information, it licenses inferences. Ask what specific wrong prediction the nickname generates. That is the question that separates students who understood §21.9 from students who highlighted it.