N1 2024-07 — audit report

66 questions (問題1–13, the non-listening half). Generated from the OCR and correction audit trails.

OCR confidence

Source: booklet-scan — quality 4/5: Clean, high-contrast grayscale scan of the original printed booklets; body text, boxed question numbers, furigana, and the 問題13 table are all crisply legible. Every page carries a light diagonal Chinese watermark (贵阳桃子日语) plus a 小红书 ID: 4675123449 stamp in the bottom-right corner, overlapping text on most pages without impairing legibility. Minor issues: slivers of an adjacent page along the left edge of both covers, a faint scanner streak at the top of page 45, and occasional grain; not a retypeset digital text, so short of a 5. Notably this scan preserves the authentic Q8, Q36, and Q37 content that is corrupted in the retypeset N1_2024-07 source.

Every section was transcribed twice independently (a full-page pass and a double-resolution half-page pass) and reconciled. Of 13 sections, 6 agreed on all content between the two passes (reconciled automatically, differing only in transcriber notes) and 7 had at least one content difference resolved by a third adjudication pass that re-read the page images. 51 leaf-level differences were examined in total.

Residual OCR risks (21):

• 問題1: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
• 問題2: None significant: A and B agreed on every leaf, and all 7 stems plus all 28 answer options were re-verified character-by-character against the double-resolution images. Watermark strokes overlap some glyphs (e.g. Q9 answers, Q13 stem) but the text remains unambiguous at double resolution.
• 問題3: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
• 問題4: Question 23 headword 填 vs 塡 (JIS glyph variant): the scan resolution cannot definitively distinguish the variants; transcribed as 補填.
• 問題4: Question 20 option 1: the underline under 風潮 is faint on the full-page render but visible on p05-t; transcribed as underlined.
• 問題5: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
• 問題6: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
• 問題7: 「自分自身の規制、相手の規制」 reads oddly in context (規制 = regulation); both transcribers and high-zoom verification agree it is printed as 規制, but the source booklet may itself be a flawed retypeset of different original wording.
• 問題7: 「わざわざ人には会いに行こうとしたのは」 is awkward Japanese but verified as printed; transcribed verbatim per conventions.
• 問題7: A light diagonal watermark crosses some lines of the passage; no characters appeared obscured at high zoom, but very subtle punctuation under the watermark carries minor residual risk.
• 問題8: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
• 問題9: The 揚げ足取り（注3） mark sits visually under 揚げ; placed after the whole headword 揚げ足取り to match the gloss entry — defensible but a placement judgment call.
• 問題9: Light diagonal watermark and 小红书 stamp overlap some characters; verified through them at 2x but a few low-contrast strokes (e.g. tiny （注N） subscripts) rely on the upscaled render.
• 問題9: q54/q53 split across p21/p22 was confirmed via boxed numbers only; the p22 q54 block was not re-read character-by-character (no A/B disagreement there).
• 問題10: The 注 reference marks are printed below the line, so any inline placement is a convention choice; オツな（注2） follows the gloss headword and the mark centering, but a reader preferring strict left-edge placement could argue for a different split.
• 問題10: The watermark crosses some passage lines on p25; no character ambiguity was observed at double resolution, but heavily watermarked glyphs (e.g. in the final paragraph) were checked only at normal zoom plus the half-page renders.
• 問題11: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
• 問題12: The printed （注2） marker sits mid-word at a line break, so its original (pre-retypeset) linear position cannot be recovered with certainty; placement after もろもろの follows the gloss-line wording.
• 問題12: Passage paragraphs 1 and 3, where A and B fully agree, were verified against the double-resolution image only at reading zoom rather than character-by-character crops; a correlated OCR error there is unlikely but not formally excluded.
• 問題13: The speck after 以降にお渡しします。 was judged scan noise, not a typeset period; if the booklet is a flawed retypeset there is a small chance it is a printed stray "." that should have been transcribed.
• 問題13: Header-row cells 期間※/平日/土曜・日曜 appear bold gothic on shading and were transcribed as <th> by both transcribers; this is consistent with the image but the bold-vs-shading distinction is a judgment call.

Source-text quirks the OCR transcribed verbatim (printed typos/recall artifacts; the ones judged to be retype errors were corrected in the layer below) — 20 noted at OCR time:

• 問題6: Question 36 is the only question with the ★ in the second slot ( _ ★ _ _ ); 37 has it in the third; 38-40 have it in the third as well. Verified against zoomed crops.
• 問題6: Question 36 has its ★ in the second of the four slots ( _ ★ _ _ ); questions 37-40 all have it in the third slot, as printed.
• 問題6: In question 37 the slot run is followed directly by a full-width period (… _ _ ★ _ 。専門家に…), as printed.
• 問題7: First body paragraph reads 「わざわざ人には会いに行こうとしたのは」, which is slightly awkward Japanese; possibly a retypeset flaw, transcribed verbatim as printed.
• 問題7: Second paragraph reads 「自分自身の規制、相手の規制にもつながる」 — 規制 both times, verified at high zoom (possibly the original text uses different wording, but this is what is printed).
• 問題8: Question 48's options are noun phrases ending in 人 with no closing 。, as printed.
• 問題8: None of the four passages prints a source citation, which is unusual for 問題8; the retypeset source appears to omit them.
• 問題9: Layout: passage (1) on p17 with q49-50 on p18; passage (2) on p19 with q51-52 on p20; passage (3) and q53 on p21, q54 alone on p22; passage (4) on p23 with q55-56 on p24. All boxed question numbers 49-56 verified against the images. The double-resolution -t halves carry all section content; the -b halves of p18/p20/p22/p24 are blank except footers.
• 問題9: Inline note-reference marks （注N） are printed as tiny subscripts beneath the glossed term and ARE transcribed (transcriber B omitted them; the marks are genuinely on the page, verified at 2x). Placement: for plain noun/adverb headwords the mark follows the whole word (ドラマ（注1）, 筋立て（注2）, おぼろげながら（注3）, カブトムシ（注1）, 突然変異（注2）, うねり（注）, 論敵（注1）, 揚げ足取り（注3）, 物知り顔（注7）); for sino-noun + する/な/される inflections it follows the sino-noun stem before the inflection (淘汰（注3）された, 浅薄（注2）な, 血肉化（注4）する, 没入（注5）して, 内在化（注6）する), matching where the subscript visually sits.
• 問題9: Source spelling kept as printed, e.g. passage (3) 行なって (not 行って), verified at 2x.
• 問題9: Verified at 2x from the -t double-resolution renders: the （注N） subscript marks, all furigana, the passage-3 underline, and the q49/q50/q55/q56 stems and answer choices.
• 問題10: No source citation （～による） line is printed after the passage or after the 注 glosses; the passage ends directly with the four 注 lines. Unusual for 問題10 but verified against the print.
• 問題10: The passage prints 多少はわからないでも (not わからなくても) and わからないのを飼っておける (no もの); both verified at high zoom.
• 問題11: Passage B, first paragraph: the quoted advice ends with a period inside the brackets, 「無理をしないであきらめたほうがよい。」, confirmed at high zoom; transcribed as printed.
• 問題11: Passage A uses 一所懸命 (not 一生懸命); confirmed at high zoom and transcribed as printed.
• 問題12: The phrase 「反抗は望ましいから、大いに結構だとも違う。」 reads oddly but is clearly printed that way (re-verified at high zoom; possibly a retypeset flaw); transcribed verbatim.
• 問題12: The passage mixes 思うとおり and 思いどおり; both forms are as printed.
• 問題13: Question numbers 65 and 66 as printed match the expectation.
• 問題13: 開館時間 table: the shaded header row has an empty first cell over the library-name column. 中央図書館 and 文学部図書館 each vertically span their two period rows; verified against the double-resolution scan that these cells are unshaded and in regular type, so they are <td rowspan="2"> cells, not <th> (rowspan used since the allowed tag set has no other way to express merged cells).
• 問題13: The breadcrumb line 秋川大学＞秋川大学図書館＞一般利用 uses full-width ＞ separators as printed; it is set in gothic type but lighter than the bold title, so it is not marked <b>.

Corrections applied (2)

Evidence-based reconstructions of the original exam text. This section is the canonical correction log for the sitting.

• 問9 q0 passage: 「（注4）するこ」 → 「する（注4）こ」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問9 q0 passage: 「（注6）する。」 → 「する（注6）。」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]