Copy Editing & Proofreading Intern

Copy editing is the last defence before publication, and it demands reading in a way ordinary reading does not: slowly, suspiciously, and against the clock.

This task is entirely desk based and uses material you prepare yourself.

First, create your material.

Take three published news stories and rewrite each one badly, deliberately introducing a mix of problems: grammatical errors, inconsistent spellings of the same name, factual contradictions within the piece, buried leads, wandering structure, unattributed claims, and repetition.

Keep a private answer key recording every flaw you planted.

Preparing material this way teaches you to see errors from both sides.

Leave the drafts for a day so you read them fresh.

Now edit each one properly.

Work through several passes rather than one, since trying to catch everything at once catches nothing.

Read first for sense and structure, then for accuracy and attribution, then for grammar and punctuation, then for consistency of names, numbers, and style, and finally read the piece aloud to catch what the eye skips.

Mark every change visibly using tracked changes or a clear marking system, and write a brief comment explaining each substantive edit rather than only correcting silently.

Then compare your edits against your answer key and record which planted errors you missed, since the misses are the most useful part of this exercise.

Edit with restraint as well as thoroughness.

A copy editor's job is to make the writing correct and clear, not to rewrite it in their own voice.

Where you changed something that was merely different from how you would have written it, rather than actually wrong, reverse the change and note it.

Learning this restraint is as important as learning to spot errors.

Then time yourself on the third draft.

Real copy editing happens under deadline, and accuracy under time pressure is a different skill from accuracy with unlimited time.

Record how long you took and which error types you began missing as the clock ran.

Finally, write a note of about 200 words on what you missed, why you think you missed it, and what change to your process would have caught it.

Submit the flawed drafts, your edited versions with changes marked, the answer key, the restraint reversals, the timing record, and the note.

Every newsroom runs on a style guide, which exists so that a publication reads as one voice rather than many.

Applying one consistently is less about memory than about building the habit of checking.

This task is entirely research and desk work.

First, study a real style guide.

Several major news organisations publish theirs openly.

Read the sections covering numbers, dates and times, titles and honorifics, place names, abbreviations and acronyms, capitalisation, and quotations.

Build yourself a quick reference sheet covering the thirty rules you expect to use most often.

Now apply it.

Take five pieces of news writing on different subjects, ideally from publications following different conventions so that genuine inconsistencies exist.

Edit all five to conform fully to your chosen guide.

Maintain a change log recording every edit: the original text, the corrected text, and the specific rule applied.

This log is the real output, because it proves the edits follow the guide rather than your preference.

Next, identify the hard cases.

Every style guide has gaps, and you will hit situations it does not cover.

Record at least five such cases, state the decision you made, and justify it by analogy to the closest rule that does exist.

Finally, write a short note proposing three additions to the guide that would resolve the ambiguities you encountered.

Pay particular attention to the rules that exist for accuracy rather than consistency.

Guidance on how to describe people, how to report allegations before a court has ruled, and how to handle terms around identity and community are not stylistic preferences but protections against misrepresentation.

Record which rules in your guide fall into this category and why each exists.

Then examine where style rules conflict with clarity.

Occasionally the prescribed form is more awkward than the alternative, and a copy editor has to decide.

Identify two such instances in your editing and explain which you prioritised.

Finally, write a short reflection on the purpose of a style guide, addressing why consistency matters to a reader who never consciously notices it.

Submit the reference sheet, the five edited pieces, the complete change log, the accuracy rules analysis, the hard case decisions, the conflicts, and the proposed additions.

Headline writing is the most constrained writing in journalism: it must be accurate, clear, and appealing, all within a character count, and it is read by far more people than the story beneath it.

This task is fully desk based.

Select six news stories covering different subjects and read each one completely, since the most common headline failure is headlining a story the writer has only skimmed.

For each story, write a headline within a strict limit of 70 characters including spaces.

Then write a standfirst of 25 to 30 words that adds genuinely new information rather than restating the headline in longer form.

For every headline, record the character count and write one line on what you chose to prioritise and what you had to leave out.

Now apply the accuracy test, which is the core of this task.

For each headline, ask whether a reader who read only the headline would come away with an accurate impression.

Identify any headline that overstates certainty, implies causation the story only suggests, or attributes a claim as fact.

Rewrite every one that fails, and keep both versions.

Next, compare your work against the headline each publication actually used.

Write a short assessment of the differences and be honest about which is better.

Finally, write a half page note on the three specific ways headlines most often mislead, drawing on examples from your own drafts.

Then write each headline a second way under a tighter limit of 45 characters, which is roughly what a mobile screen or a push notification allows.

Severe constraint is the real test of whether you have identified the core of the story, and the shorter version frequently turns out to be the better one.

Also write the headline for someone who will see nothing else.

Many readers encounter only a headline in a feed or notification and never open the story, which means a headline that is technically accurate but incomprehensible alone has failed them.

Check each of yours against this test.

Finally, write a half page note on the three specific ways headlines most often mislead, drawing on examples from your own drafts, and explain the one rule you would apply to every headline you ever write.

Submit all headlines and standfirsts with counts, the shortened versions, the rewrites, the comparisons, and the note.

Corrections are published evidence of where editing processes fail.

Studying them systematically shows which errors recur and at which stage they should have been stopped.

This is a research and design task done entirely online.

Collect at least ten published corrections from news organisations.

Many publish these on dedicated corrections pages, and they are freely available.

Record each one with the publication, date, what was wrong, and what the correct position was.

Now classify them.

Build categories covering factual errors such as wrong names, dates, or figures; attribution errors where a quote or claim was wrongly assigned; numerical errors including arithmetic and unit mistakes; headline errors where the headline misrepresented accurate copy; and context errors where true facts created a false impression.

For each correction, identify the stage at which it should have been caught: the reporter's own checking, the copy desk, the fact checking stage, or the headline writer.

Analyse the pattern.

Identify which category recurs most, which stage fails most often, and which errors are most damaging to reader trust even when factually minor.

Now design the preventive process.

Write a correction policy covering how errors are reported, who decides, how quickly corrections are published, how prominently they appear, and how the record is kept.

Add a pre publication checklist targeting specifically the error types your audit found most common.

Examine the wording of the corrections themselves, which varies enormously.

Some state plainly what was wrong and what is correct.

Others are so hedged that a reader cannot tell what the error was.

Collect examples of both and write about why clear corrections serve a publication better than evasive ones, even though they are more uncomfortable to publish.

Then consider what never gets corrected.

Misleading framing, an inaccurate headline over an accurate story, and missing context rarely produce formal corrections even though they mislead more readers than a wrong date does.

Write a section on how your policy would handle these.

Finally, write model correction wording for three of the errors in your collection, in the exact form you would publish them.

Submit the correction collection with sources, the classification table, the pattern analysis, the wording study, the policy, the model corrections, and the checklist.