Explanatory package problems: from a vague scope to stale updates

What to take away

Explanatory package problems can show up in a package whose every sentence is grammatical and whose every component looks finished. They usually start with structure rather than prose: the package promises no clear answer, its parts repeat one another, or evidence loses status as it moves from source notes into timelines, cards, charts, and summaries. These failures appear in coverage of stories already in the news — a ruling, a data release, a disaster, a policy change — where readers come for what led here, who the parties are, and which parts remain genuinely uncertain.

Problem 1: the assignment is only a topic

Symptom: The plan says "housing," "inflation," or "the court case." Reporters collect related facts without knowing what the package must explain, and the finished pieces never answer the question their headlines raise.

Fix: Write a question, audience, boundary, and main provisional answer. Name what new evidence could change the answer. Reject material that does not help complete that job. The package contract comes from building an explanatory package.

Problem 2: every component retells the main article

Symptom: The timeline repeats paragraphs with dates attached. The glossary restates the lead. Profiles recount the same conflict without explaining authority.

Fix: Give every component a different reader task and required evidence. Merge pieces with the same task. A timeline should explain consequential sequence; a profile should explain a person's role and relevant actions. Task-by-task evidence demands are tabulated in the context elements comparison.

Problem 3: background becomes a fact dump

Symptom: Old events appear because they are interesting, not because they explain the present.

Fix: Add a "why this matters now" field to every proposed context item. Remove entries with no defensible answer. Summarize long periods when individual dates add no meaning. Selection by function is the core rule of the background briefing guide.

Problem 4: a visual carries no independent context

Symptom: A chart title says "Cases over time," while the unit, population, period, revisions, and source sit elsewhere or nowhere.

Fix: Make the visual interpretable on its own. The CDC's collection of visual communication resources advises choosing visuals that support the main message and giving them clear titles, labels, and captions. For news work, add the dataset, date, definitions, and qualifications needed for the specific claim.

A visual should not be chosen because data exist. First write the sentence it is meant to support. Then check whether the measure answers that question. If not, find different data or change the sentence.

Problem 5: source status disappears

Symptom: A preliminary estimate enters the main story with a caveat, then appears as a settled number in a chart card and FAQ.

Fix: Put status in the shared claim record and in every display field. Search all package components when the source is revised. A correction is incomplete if only one occurrence changes.

Problem 6: chronology turns into causation

Symptom: Two events appear next to each other and the package implies that the first caused the second.

Fix: State what the chronology proves: order. Add separate evidence for mechanism and alternatives before using causal language. If the evidence establishes only association, say so. The restrained-wording list in the news timeline method supplies the verbs.

Problem 7: a supplied record is treated as neutral

Symptom: A campaign, company, agency, or advocacy group provides a timeline, map, or document set, and the package adopts its selection without examining omissions or purpose.

Fix: Identify the creator, audience, purpose, date, and missing material. The National Archives worksheet on understanding perspective in primary sources asks readers to consider background, context, the creator's perspective, and how that perspective affects the source. Perspective does not make a record useless, but it changes what the record can establish.

Compare the supplied selection with an independent record index, docket, archive, or dataset. Describe meaningful omissions instead of silently filling them with assumptions.

Problem 8: false balance replaces evidentiary weight

Symptom: Two opposed quotations receive equal space even though one side cites a current record and the other offers an unsupported denial.

Fix: Report relevant positions, then show what each rests on. Fairness requires accurate representation and a chance to respond. It does not require treating unequal evidence as equal. Weighing evidence over quote counts is a fix shared with attribution reporting problems.

Problem 9: the package mixes time states

Symptom: A completed vote, pending appeal, old rule, and proposed amendment are written in the same tense.

Fix: Add exact dates and status labels. Identify which rule controls now, when a change could take effect, and what remains contingent.

Problem 10: maintenance has no owner

Symptom: The main article updates after a new report, but sidebars and FAQs remain stale.

Fix: Assign one package editor, keep a component inventory, and define update triggers. Each shared claim should list every place it appears. Retire a component when it can no longer be maintained reliably. The maintenance section of the background briefing editorial checklist assigns those owners and triggers.

A short recovery sequence

When a package built around a story already in the news is in trouble, work in this order:

Common questions

Is repetition always a problem?

No. A short orienting sentence may be needed in several entry points. Repeating full explanations or evidence without a new reader job is wasteful.

Can a partisan source be used?

Yes, for what that source said, did, filed, or recorded within its competence. Check factual claims independently and disclose relevant perspective.

Should a weak chart be replaced with prose?

Often. If the data cannot support a clear comparison, prose that states the limit may be more accurate than a polished display.

When does stale context require a correction?

When it makes a current factual statement wrong or materially misleads readers. Minor maintenance can be logged as an update under the newsroom's policy.

Do these problems get worse in breaking news?

Yes. In the first hours the package contract is often unwritten, so scope, repetition, chronology, and source status drift fastest. Fixing the reader question early, and recording the status of each claim as it arrives, is the cheapest correction available.