The Cross-Reference That Points Nowhere
A 60-page policy manual uses numbered headings (1, 1.1, 1.1.1) generated by the heading styles. The body text contains 38 cross-references of the form "see section 4.2.1", all typed by hand as plain text. During review, section 4.2 is deleted and a new subsection is inserted at 2.3, which renumbers everything after it. The reviewer also asks you to add a table of contents and a list of figures. The document must go to a regulator as a PDF next week, and a second round of edits is expected before then.
1. Explain why the 38 hand-typed references are now a correctness risk and what should replace them.
Read the worked answer
Hand-typed references are static text that does not track numbering. When 4.2 was deleted and 2.3 inserted, every subsequent heading number changed, but the 38 typed strings still say what was true at typing time. So the document now contains references that point to the wrong section - or to a section that no longer exists - and nothing flags the error.
The fix is to replace them with automatic cross-references to the heading's number, inserted through the cross-reference feature so each reference stores a link to the target heading. When headings are renumbered, the references update with the field update. This converts a manual proofreading problem (38 items, each needing human verification) into a single update operation.
This matters especially because a second edit round is coming: any manual approach would have to be redone, whereas automatic references survive the next renumbering.
2. The table of contents and list of figures will be generated. What must be true of the document before those are trustworthy, and what is the risk of generating them now?
Read the worked answer
Before a TOC or list of figures is trustworthy, the document must have consistent use of the built-in heading styles and caption styles - the TOC is built from heading levels, and the list of figures is built from captions. If any chapter title was formatted manually to look like a heading (bold, larger font) rather than tagged with Heading 1, it will silently be omitted from the TOC.
The risk of generating them now is premature generation: the fields capture the current state, and if the second edit round changes headings or captions, the TOC and list of figures go stale unless the fields are updated (and page numbers recalculated) before export. The safe sequence is: normalise styles → generate fields → update all fields → export to PDF, and repeat the update step after the second edit round.
3. Should the PDF be produced by printing to PDF or by exporting? Justify in terms of the references and navigation.
Read the worked answer
Export, not print-to-PDF.
Export preserves the document's structural metadata: heading bookmarks become PDF navigation, and cross-references and TOC entries can be exported as live internal hyperlinks. Print-to-PDF flattens the result to a picture of pages - the links become inert, and the regulator cannot click "see section 4.2.1" to jump there.
Caveat: export only helps if the fields were updated first. An export of stale fields produces a beautifully navigable PDF pointing at wrong numbers, which is worse than an obviously broken one.