Instructor Notes — Chapter 40

Your Peptide Evidence Dossier: Putting It All Together Part 7 · capstone of the method arc · prerequisites: Chapters 5 and 37

What this session is

This is the graduation exercise, and it is a workshop, not a lecture. Nearly everything of value in Chapter 40 happens when a student opens their own document and finds something wrong in it. If you teach this chapter as content — twelve fields, four checks, two biases — students will nod, agree, and leave with an unaudited dossier. Budget for making them work in the room.

The single most important framing to establish in the first five minutes: the audit is not about whether their ratings are correct. Students arrive expecting to be graded against Appendix A. They did that in Chapter 37. This session asks whether they applied one standard, which is a question their own file can answer without any external reference. Getting that distinction landed early determines whether the rest of the session works.

Do not let this feel like the last class. It is not. Part VIII follows and §40.10 is a handoff. If your course ends here, say so explicitly and separately — but do not let students infer it from the chapter, because the chapter is written as the end of a training rather than the end of a book.

Time allocation (a 3-hour block, or two 90-minute sessions)

If you have only 90 minutes: cut the transfer exercise to homework, run Case Study 1 as the anchor, and protect the four-checks segment at all costs. It is the section the chapter is built around.

Prerequisites students actually need

Chapter 5 (rating system, six rules, the two kinds of ❌) and Chapter 37 (master table) are the declared prerequisites. In practice, students who did not do Chapter 1's Step 3 — write down what you already believe, with a confidence number, and date it — cannot do §40.4 properly. Check for this in advance. If a cohort skipped it, substitute: have them write their current beliefs first, then compare against Chapter 37, and be explicit that the exercise is weaker than it should be because the "before" measurement is missing.

Appendix C should be open on every desk. This chapter is the assembly instruction for that appendix, and §40.2 deliberately does not reproduce its table.

Where students reliably struggle

They fill gaps before making the completion map. Exercise C requires the map first. Students who fill as they go lose the by-field tally in D, which is the finding — that 9, 10, and 12 are empty across the board — and it lands much harder as their own measurement than as your assertion. Enforce the ordering.

They write "no known side effects" in Field 9. Every cohort does this. The correction is not a matter of tone: for a compound with no human trials and no reporting pathway, that sentence describes a surveillance system that does not exist. Have one student read theirs aloud, then walk the five links of the reporting chain (from the Case Study 2 marking guidance) and ask which exist.

They cannot find a same-evidence pair. Almost always because they are looking among their well-evidenced entries. Redirect them to the entries with no human evidence, where nothing external constrains the rating and non-evidence reasons operate freely. That is where the pair usually is.

They resist the Field 12 check on their ❌ entries. This is the reliable friction point of the session and it is worth having a prepared response. The objection is always some version of "my ❌ is right, so specifying what would move it is bureaucracy." Case Study 1 exists to answer it: Entry 2's ❌ is correct and was produced by distaste, and the same method produced ⚠️ on identical evidence one entry later. Being right for bad reasons feels identical to being right.

They treat the harsh bias as the responsible one. Expect this, especially from science-track students. See Discussion Prompt 3.

They confuse "nobody knows" with "promising." §40.7's three sentences usually need to be worked aloud, not read. The counterintuitive claim — that a compound with a large null trial is better understood than one with no trials — is the one to spend time on.

Sensitive handling

Some students in any cohort are currently taking one of the compounds in their dossier, or have a family member who is. §40.4 asks them to discover that they were systematically generous toward exactly those compounds, which is uncomfortable in a way the chapter intends and a classroom can amplify.

Two guardrails. First, make the §40.4 direction exercise individual and unshared by default — invite volunteers, never call on someone. Second, hold the line the chapter holds: a rating is an input to a decision, not the decision, and two people with identical dossiers can reasonably choose differently. A student discovering a generous bias has not discovered that their choice was wrong.

Never let the session turn into advice about anyone's actual use. Redirect to §40.7's verdict-to-question conversion: the output is a better question for a clinician, not a recommendation from you or from the room.

Assessment suggestions

• Best single artifact: the student's audited dossier, submitted with the pre-audit frozen version alongside. Grade the delta, not the document. What changed, and can they say why?
• Best single written item: Exercise AG † — a complete Field 6 row and a Field 12 for a non-peptide claim. It tests transfer, specificity, and the reasonable-person test in one page.
• Best discussion artifact: Exercise AI † — attempting Field 5 on a societal claim. Collect these; they are the pre-test for Part VIII and make an excellent opening for Chapter 41.
• Do not grade the ratings. Grade whether the standard is visible and consistent.