Editing Accuracy | Avoid False Positives and Unnecessary Changes

Editing accuracy is not the ability to change a large number of words. It is the ability to change the right words and leave the right words alone.

That sounds obvious until a learner enters a supplied-text editing task with a dangerous assumption: there must be something wrong here, so I should keep searching until I change something. From that moment, correct grammar becomes vulnerable. A legitimate tense shift starts to look inconsistent. A linking-verb adjective looks as though it needs -ly. A collective noun is forced into one agreement pattern even when the target standard accepts another. A passive sentence is rewritten simply because active voice feels more direct. A correct preposition is replaced because another preposition is familiar. The learner is no longer editing errors. The learner is manufacturing them.

This article is the canonical SETC longform for editing accuracy, false-positive control and unnecessary-change prevention. It belongs to the Secondary English Editing Examination Performance lane and builds on the previous spokes:

• English Error Correction Exercises With Answers — detection before repair;
• Grammar Editing Exercises — deep control of major grammar families;
• Passage Editing Practice — context and meaning preservation;
• English Editing Under Time Pressure — scan, verify and stop;
• Sentence Correction Exercises — controller-first diagnosis;
• Editing Error Types — personal grammar error mapping.

This page has one narrow job: teach the learner to preserve correct language until there is enough evidence to justify a change.

The operating model is:

SUSPECT → IDENTIFY THE RELATIONSHIP → LOOK FOR REAL EVIDENCE → TEST THE ORIGINAL → CHANGE ONLY IF THE ORIGINAL FAILS → REPAIR MINIMALLY → VERIFY ONCE → STOP.

Maren will ask whether a structural relationship genuinely fails. Iona will demand evidence before accepting a correction. Leonie will protect the learner from time-consuming second-guessing and late answer changes. Their shared aim is precision.

1. What is a false positive in editing?

A false positive occurs when the learner identifies correct language as erroneous and changes it.

Example:

A number of students are waiting outside.

A learner remembers that number is singular and changes are → is. The learner has applied a real grammar idea to the wrong structure. The sentence was correct. The edit creates an error.

False positives matter because they reveal something different from ordinary mistakes. The learner may know many rules. The problem is not ignorance. The problem is miscalibration: the evidence threshold for changing language is too low.

2. Accuracy has two sides

Editing accuracy requires both:

• sensitivity: find genuine errors;
• specificity: preserve genuine correct forms.

A learner who never changes anything has high preservation but poor detection. A learner who changes half the passage may detect many real errors but destroy correct language. Strong editing balances both.

3. Precision and recall

Two useful concepts:

• Recall: of all genuine errors present, how many did you find?
• Precision: of all edits you made, how many were actually justified?

A strong editor wants both high recall and high precision.

4. Why advanced learners sometimes over-edit

As learners know more grammar, they can imagine more alternatives. A beginner may not see any option. An advanced learner may see four possible rewrites. That larger hypothesis space creates a new risk: confusing possible with necessary.

Editing asks a different question:

Is the original unacceptable under the target standard, or merely different from the version I would personally write?

5. The evidence threshold

Before changing a form, require at least one specific piece of evidence:

• subject and verb disagree;
• auxiliary requires a different verb form;
• time expression conflicts with tense;
• grammatical slot requires another word class;
• lexical frame selects another preposition or complement;
• pronoun case or reference fails;
• comparison is structurally illogical;
• modifier attaches to the wrong subject;
• coordination is not parallel;
• sentence boundary is broken.

“It sounds unusual” is not enough.

6. The burden of proof belongs to the editor

The supplied sentence existed before you touched it. Your correction changes the text. Therefore the burden of proof belongs to you.

A useful training sentence:

If I cannot name the relationship that fails, I do not yet have enough evidence to change the form.

7. Unfamiliar does not mean incorrect

Correct English includes structures learners may encounter less often:

• Only after the second test did the fault become clear.
• Had the warning arrived earlier, the team might have changed course.
• What the committee needs is more evidence.
• Not only did the method save time, but it also reduced waste.
• Who did you speak to?
• Everyone should bring their notes.

If unfamiliarity automatically triggers correction, advanced structures become false-positive traps.

8. Different does not mean incorrect

Two expressions can both be grammatical:

• Before the teacher arrived, the students began the warm-up.
• Before the teacher arrived, the students had begun the warm-up.

The second emphasizes earlier completion more strongly; the first can be perfectly adequate because before already establishes sequence.

Do not replace one correct option with another unless the task asks for style revision.

9. Better does not mean necessary

An editor may prefer:

The researchers analysed the results carefully.

over:

The results were carefully analysed by the researchers.

But the passive sentence is grammatically correct. A supplied-text grammar task should not be turned into a personal style workshop.

10. The minimal-change principle

When a genuine error is found, change the smallest unit that repairs it.

Original:

The list of names are complete.

Minimal:

The list of names is complete.

Unnecessary rewrite:

All the names have been completely listed.

The rewrite changes meaning, structure and vocabulary. The one-word repair proves the diagnosis more clearly.

11. The unnecessary-change test

After proposing an edit, ask:

1. What exact error does this edit repair?
2. Would the original still be grammatical if I left it alone?
3. Does the edit alter meaning, emphasis, register or reference?
4. Can I repair the same problem with a smaller change?

12. Correct-line preservation is an active skill

Leaving a line unchanged is not “doing nothing.” It is a decision supported by analysis.

Strong editing sometimes concludes:

I inspected the relationship. It survives. No change.

13. The three possible outcomes

Every sentence should end in one of three states:

• Change: specific grammatical failure proven.
• Keep: original survives the relevant checks.
• Return: uncertainty remains and later context or time may help.

There is no fourth category called “change because I feel nervous.”

14. False-positive family 1 — Agreement overcorrection

Common traps:

• A number of students are…
• The number of students is…
• collective nouns;
• relative-clause verbs;
• plural scientific data;
• mass-noun singular agreement.

15. Agreement Control Set A

All sentences are correct in the intended standard. Do not change them.

1. A number of students are waiting outside.
2. The number of students is increasing.
3. The series of workshops begins next month.
4. The equipment was delivered yesterday.
5. The news was unexpected.
6. The students who study regularly improve.
7. One of the students who work in the laboratory has published a report.
8. The team is meeting this afternoon.
9. The team are arguing among themselves.
10. The data are consistent with the model.

Items 8–10 may depend on variety or target convention. The key skill is not to force a change before identifying the expected standard.

16. Agreement Control Set B

Some are wrong. Change only those that fail.

1. The quality of the samples has improved.
2. The quality of the samples have improved.
3. A number of visitors are waiting.
4. A number of visitors is waiting.
5. The number of visitors is rising.
6. The number of visitors are rising.
7. Each of the reports contains a summary.
8. Each of the reports contain a summary.
9. The photographs in the archive have faded.
10. The photographs in the archive has faded.

Key

2, 4, 6, 8 and 10 require change. The others should be preserved.

17. False-positive family 2 — Tense overcorrection

Tense is especially vulnerable because multiple grammatical forms can express different aspects of time.

Incorrect editing often follows simplistic rules:

• “Keep all verbs in one tense.”
• “Use past perfect whenever two past events appear.”
• “Use present perfect with every since sentence regardless of meaning.”

These shortcuts create false positives.

18. Tense Control Set A

All are correct.

1. The museum opened in 1998 and has expanded several times since then.
2. Last year, the team tested the prototype. It now uses the same method in two other sites.
3. Before the teacher arrived, the students began the warm-up.
4. Before the teacher arrived, the students had begun the warm-up.
5. The train leaves at six tomorrow.
6. When the meeting starts, we will review the results.
7. The researchers have monitored the site since 2021.
8. Yesterday, the researchers published an interim report.
9. Water expands when it freezes.
10. By the time Leonie arrived, the presentation had already started.

19. Tense Control Set B

Correct only genuine failures.

1. She submitted the report yesterday.
2. She has submitted the report yesterday.
3. We have lived here since 2022.
4. We live here since 2022.
5. When the bell rings, the students will leave.
6. When the bell will ring, the students will leave.
7. By the time I arrived, they had left.
8. By the time I arrived, they have left.
9. The project began in January and is still continuing.
10. The project has begun in January and is still continuing.

Key

2, 4, 6, 8 and 10 require repair in the intended readings.

20. False-positive family 3 — Linking-verb adjectives

Learners who overlearn “verbs need adverbs” may change correct adjectives after linking verbs.

Correct:

• The plan seems sensible.
• The fabric feels soft.
• The soup tastes good.
• The speaker sounded confident.

The adjective describes the subject through the linking verb.

21. Linking-Verb Control Set

1. The plan seems sensible.
2. The plan seems sensibly.
3. The machine operates smoothly.
4. The machine operates smooth.
5. The fabric feels soft.
6. The fabric feels softly.
7. The speaker sounded confident.
8. The speaker sounded confidently.
9. Leonie looked carefully at the graph.
10. Leonie looked careful at the graph.

Key

1, 3, 5, 7 and 9 are correct. Preserve them. Repair 2, 4, 6, 8 and 10 according to the intended meanings.

22. False-positive family 4 — Passive voice

Passive voice is grammatical and often useful. Do not change it merely because active voice is sometimes stylistically preferred.

Correct:

• The samples were stored at low temperature.
• The bridge was built in 1980.
• The report was completed yesterday.
• The results were checked by two analysts.

23. Passive Control Set

Mark C or E.

1. The samples were stored safely.
2. The samples were store safely.
3. The bridge was built in 1980.
4. The bridge was build in 1980.
5. The report was completed yesterday.
6. The report was complete yesterday.
7. The figures were checked by Iona.
8. The figures were check by Iona.
9. The equipment was damaged during transport.
10. The equipment were damaged during transport.

Key

1 C. 2 E. 3 C. 4 E. 5 C. 6 can be correct if complete is an adjective meaning finished; context decides. 7 C. 8 E. 9 C. 10 E.

24. False-positive family 5 — Formal inversion

Correct formal English sometimes places the auxiliary before the subject:

• Only then did I understand.
• Never had the researchers seen such a pattern.
• Not only did the method save time, but it also reduced waste.
• Had the warning arrived earlier, the team would have changed course.

Do not flatten these structures into ordinary declaratives simply because they look unfamiliar.

25. Inversion Control Set

1. Only after the second test did the fault become clear.
2. Only after the second test the fault became clear.
3. Never had Maren seen such a result.
4. Never Maren had seen such a result.
5. Not only did Leonie finish on time, but she also checked her work.
6. Not only Leonie finished on time, but she also checked her work.
7. Had the warning arrived earlier, the team might have changed course.
8. Had the warning arrived earlier, the team might changed course.

Key

Odd-numbered items are correct. Even-numbered items contain structural failures.

26. False-positive family 6 — Sentence-final prepositions

In standard modern English, sentence-final prepositions can be grammatical:

• Who did you speak to?
• This is the topic we were talking about.
• Which chair did you sit on?

Formal alternatives also exist:

• To whom did you speak?
• This is the topic about which we were talking.

Do not automatically “correct” one into the other unless register requirements justify it.

27. False-positive family 7 — Singular they

Singular they is widely used in modern standard English and is especially useful when gender is unknown, irrelevant or nonbinary.

Example:

Every student should bring their notes.

Do not replace their automatically with his or her unless the target standard specifically requires an older convention.

28. False-positive family 8 — Collective nouns

Collective-noun agreement varies by meaning and variety.

Examples:

• The team is meeting today. — group as one unit.
• The team are arguing among themselves. — members emphasized, common in British English.

Know the target assessment convention before deciding. Do not convert recognized variation into “error.”

29. False-positive family 9 — Data

Data may take plural agreement in formal scientific usage:

The data are convincing.

In much contemporary general English, singular mass treatment also appears:

The data is convincing.

The correct examination decision depends on target standard. A global editing article should teach the variation rather than pretend only one form exists.

30. False-positive family 10 — Before/after with simple past

Sequence words can make time order clear without past perfect.

Correct:

After the meeting ended, we left.

Also correct:

After the meeting had ended, we left.

The second highlights prior completion more explicitly. Do not force it when the first already works.

31. False-positive family 11 — Stative versus dynamic meaning

Some verbs change meaning depending on aspect:

• I think you are right.
• I am thinking about the problem.
• She has a car.
• She is having lunch.

Do not apply “stative verbs never use continuous” mechanically.

32. False-positive family 12 — Articles in generic meaning

Correct generic patterns include:

• Books can change lives.
• Water expands when it freezes.
• The tiger is an endangered animal. — species-level generic in one pattern.
• A tiger is a powerful predator. — representative singular.

Article use is meaning-dependent. Do not insert the everywhere.

33. False-positive family 13 — Repetition for precision

In technical writing, repeated nouns may be better than varied synonyms.

Example:

The sensor recorded temperature. The sensor also measured humidity.

A learner may want to replace the second sensor with device for variety. But repetition may preserve precision. Editing is not thesaurus substitution.

34. False-positive family 14 — Accepted preposition variation

Some structures permit more than one standard preposition depending on variety and meaning.

Examples include:

• compare with / compare to;
• different from / to / than;
• pleased with / pleased about / pleased by in different contexts;
• comment on / comment about.

Do not force a single memorized option when the original is standard and meaning-appropriate.

35. False-positive family 15 — Redundancy versus grammar

Some expressions may be stylistically redundant without being strictly ungrammatical.

The tests still remain useful.

Many editors prefer remain useful because still and remain overlap. But in a grammar-only task, this may not be a target error. Know the scope.

36. Grammar scope versus style scope

Before editing, know what the task tests:

• grammar only;
• grammar + spelling;
• grammar + punctuation;
• formal style;
• clarity;
• whole-sentence revision.

False positives often arise when the learner edits outside the stated scope.

37. Scope Control Drill

Assume a task tests grammar only. Which sentences must change?

1. The report was written by the team.
2. The team wrote the report.
3. The report was very clear.
4. The report was extremely clear.
5. The list of names are complete.
6. The students discussed the issue.
7. The students had a discussion about the issue.
8. Who did you speak to?
9. The equipment were damaged.
10. The sentence is slightly wordy but grammatically correct.

Key

5 requires agreement repair. 9 requires were → was under standard mass-noun agreement. The others may differ in style, concision or register but are not necessarily grammatical errors.

38. The confidence problem

Low confidence can trigger unnecessary changes. A learner sees a structure, feels uncertain and “fixes” it to something familiar. This is risk substitution: uncertainty about the original is exchanged for confidence in a rewrite, even when the original was correct.

39. Confidence is not evidence

High confidence can also be wrong. The learner may confidently believe:

• every data must be plural;
• every collective noun must be singular;
• every past sequence needs past perfect;
• every sentence-final preposition is wrong.

That is why the evidence threshold must be structural, not emotional.

40. Four confidence states

Confidence	Correctness	Interpretation
High	Correct	Stable knowledge
High	Wrong	Misconception
Low	Correct	Fragile knowledge
Low	Wrong	Weak knowledge

False-positive work often targets high-confidence misconceptions and low-confidence correct forms.

41. The second-guessing loop

A common sequence:

ORIGINAL LOOKS FINE → LEARNER RE-READS → UNCERTAINTY RISES → ALTERNATIVE SOUNDS FAMILIAR → ORIGINAL IS CHANGED → CORRECT ANSWER BECOMES WRONG.

The solution is not “never review.” It is “change only when review finds new evidence.”

42. Answer-changing discipline

Change a previously chosen answer when:

• you identify the actual controller;
• you notice a missed time marker;
• later context resolves the antecedent;
• you retrieve the exact lexical frame;
• you discover that the first repair created another error;
• you realize the task instruction was misread.

Do not change merely because time remains.

43. The new-evidence rule

NO NEW EVIDENCE, NO NEW ANSWER.

This single rule can dramatically reduce late false positives.

44. The one-return rule

For uncertain items, return once with a specific question. If no new evidence appears, preserve the best-supported answer and move.

Repeatedly reopening the same item often adds anxiety rather than information.

45. Verification is not reopening

After a correction, verify once:

• controller agrees;
• form fits slot;
• meaning preserved;
• no new error created.

Verification should close the decision unless contradiction appears.

46. Minimal repair reduces false positives

Every additional change is another opportunity to be wrong. Therefore minimal repair is not merely elegant; it is risk control.

47. Change-count discipline

During training, record:

• number of edits made;
• number justified;
• number unnecessary;
• number that created new errors.

This makes over-editing visible.

48. False-positive rate

False-positive rate = correct items changed ÷ correct items presented.

Track it by grammar family.

49. Precision rate

Editing precision = justified edits ÷ all edits made.

Example:

A learner makes 12 edits. Nine are justified. Precision = 75%.

That learner may look active, but one quarter of the changes damage correct language.

50. The preservation score

Another useful measure:

Preservation score = correct items left correct ÷ correct items presented.

This rewards restraint explicitly.

51. Error detection without over-editing

The goal is not to become conservative to the point of missing genuine errors. Strong editing preserves correct lines while still finding real problems.

This balance can be trained.

52. Correct-Control Set 1 — Agreement

All are correct.

1. The quality of the results is encouraging.
2. A number of participants are ready.
3. The number of participants is increasing.
4. Each of the reports contains a summary.
5. The effects of the storm were visible.
6. The series of lectures begins tomorrow.
7. The news was unexpected.
8. The equipment was expensive.
9. One of the students who work here has finished.
10. The students who work here have finished.

53. Correct-Control Set 2 — Tense

All are correct.

1. She submitted the report yesterday.
2. She has worked here since January.
3. Before the teacher arrived, the students began the warm-up.
4. Before the teacher arrived, the students had begun the warm-up.
5. When the meeting starts, we will review the results.
6. The museum opened in 1998.
7. The museum has expanded several times since 1998.
8. Water boils at 100°C under standard atmospheric pressure.
9. By the time Iona arrived, the session had started.
10. The train leaves at six tomorrow.

54. Correct-Control Set 3 — Word form

All are correct.

1. The machine operates reliably.
2. The machine is reliable.
3. The fabric feels soft.
4. Leonie touched the fabric softly.
5. The speaker sounded confident.
6. The speaker answered confidently.
7. The explanation was remarkably clear.
8. The result changed suddenly.
9. The students were bored.
10. The lesson was boring.

55. Correct-Control Set 4 — Prepositions and complements

All are correct.

1. Iona is interested in grammar.
2. Leonie is responsible for timing.
3. Maren is capable of explaining the structure.
4. The result depends on several factors.
5. The students listened to the recording.
6. The report focuses on three issues.
7. The students participated in the workshop.
8. Iona looks forward to presenting her findings.
9. Maren decided to revise the paragraph.
10. Leonie avoided repeating the mistake.

56. Correct-Control Set 5 — Advanced structures

All are correct.

1. Only then did I understand the problem.
2. Had the warning arrived earlier, the team would have changed course.
3. What the committee needs is more evidence.
4. Not only did the route save time, but it also reduced fuel use.
5. Who did you speak to?
6. Everyone should bring their notes.
7. The data are consistent with the hypothesis.
8. The team are arguing among themselves.
9. The samples were stored at low temperature.
10. Before the meeting began, we reviewed the notes.

57. Mixed Accuracy Set A — 20 items

Mark K = keep or C = change. Change only when necessary.

1. The list of names is complete.
2. The list of names are complete.
3. A number of students are waiting.
4. The number of students are increasing.
5. The result may indicate a problem.
6. The result may indicates a problem.
7. She submitted the form yesterday.
8. She has submitted the form yesterday.
9. The machine operates reliably.
10. The machine operates reliable.
11. The fabric feels soft.
12. The fabric feels softly.
13. Iona is interested in grammar.
14. Iona is interested on grammar.
15. The students who study regularly improve.
16. The students who studies regularly improve.
17. Only after the second test did the fault become clear.
18. Only after the second test the fault became clear.
19. The report was completed yesterday.
20. The report was complete yesterday.

Key

1 K. 2 C. 3 K. 4 C. 5 K. 6 C. 7 K. 8 C. 9 K. 10 C. 11 K. 12 C in the intended linking meaning. 13 K. 14 C. 15 K. 16 C. 17 K. 18 C. 19 K. 20 may be K if complete functions as adjective meaning finished; context decides.

58. Mixed Accuracy Set B — 20 items

1. The number of errors has fallen.
2. A number of errors have been found.
3. The number of errors have fallen.
4. A number of errors has been found.
5. The students have written three pages.
6. The students have wrote three pages.
7. The samples were stored safely.
8. The samples were store safely.
9. When the bell rings, we will begin.
10. When the bell will ring, we will begin.
11. Before the teacher arrived, the students began.
12. Before the teacher arrived, the students had begun.
13. The students discussed the issue.
14. The students discussed about the issue.
15. Leonie is capable of explaining the answer.
16. Leonie is capable to explain the answer.
17. The report was based on evidence.
18. The report was based in evidence.
19. Who did you speak to?
20. To whom did you speak?

Key

1 K. 2 K. 3 C. 4 C. 5 K. 6 C. 7 K. 8 C. 9 K. 10 C. 11 K. 12 K. 13 K. 14 C. 15 K. 16 C. 17 K. 18 C. 19 K. 20 K.

59. Mixed Accuracy Set C — 20 items

1. The programme aims to reduce waste, improve efficiency and lower costs.
2. The programme aims to reduce waste, improving efficiency and lower costs.
3. This method is easier.
4. This method is more easier.
5. The salary of engineers is higher than that of teachers.
6. The salary of engineers is higher than teachers.
7. Walking through the museum, Maren noticed the paintings.
8. Walking through the museum, the paintings impressed Maren.
9. Although the road was closed, we continued.
10. Although the road was closed, but we continued.
11. The alarm sounded; everyone left.
12. The alarm sounded, everyone left.
13. Everyone should bring their notes.
14. Everyone should bring his or her notes.
15. The data are convincing.
16. The data is convincing.
17. The team is meeting at three.
18. The team are arguing among themselves.
19. The results were carefully analysed.
20. The researchers carefully analysed the results.

Key

1 K. 2 C. 3 K. 4 C. 5 K. 6 C. 7 K. 8 C. 9 K. 10 C. 11 K. 12 C in formal edited prose. 13 K. 14 K though stylistically heavier. 15 K in plural scientific usage. 16 K in common singular mass usage. 17 K. 18 K in varieties permitting plural collective agreement. 19 K. 20 K.

60. Why mixed sets must contain many correct items

If every practice sentence is wrong, the learner develops a dangerous expectation: correction is mandatory. Real editing requires discrimination.

Advanced sets should include:

• clearly wrong sentences;
• clearly correct sentences;
• correct but unfamiliar structures;
• accepted variation;
• context-dependent cases;
• near-miss pairs.

61. Error density shapes behaviour

High-density sets train repair. Low-density sets train restraint. Both are needed.

62. Low-Density Accuracy Passage 1

A community archive stores photographs, letters and maps from the surrounding area. Each item is catalogued before it is placed in storage. Since the archive introduced digital records, researchers have been able to find material more quickly. A number of volunteers help with scanning. The number of fragile documents are still increasing as new donations arrive. Staff are responsible for checking each item before use.

Key

Only one clear error in the intended standard: The number…are → is. Everything else should be preserved.

63. Low-Density Accuracy Passage 2

A research team has monitored the wetland since 2020. Every month, members visit the same sites and record water level, temperature and bird activity. The equipment was replaced last year, and the new sensors have improved data quality. The team are currently discussing their individual field notes. The results from the most recent survey looks unusual and will be checked again.

Key

If plural collective agreement is acceptable in the target standard, The team are… may be correct. The clear target is results…looks → look. Do not force a change to the collective noun without standard-specific evidence.

64. Low-Density Accuracy Passage 3

Maren joined a reading group last year. Since then, she has completed twelve books and has become more confident discussing unfamiliar texts. The group meets every Thursday. Before each meeting begins, members post one question online. At the meeting, everyone should bring their notes. The group leader sometimes summarises the discussion at the end.

No changes are required.

65. High-Density Accuracy Passage 1

A school science club meet every Thursday. Each project begin with a question that students can investigate. Since the club started last year, members completed several studies. The teacher encourages students to record observations careful. One of the most useful habits are repeating a test when a result looks unusual. Students are responsible of returning equipment after each session.

Key

meet → meets; begin → begins; completed → have completed if the continuing period is intended; careful → carefully; are → is; of → for.

66. Why high-density success does not prove accuracy

A learner can perform well when nearly every sentence is wrong because expectation helps detection. Low-density and correct-control tasks are necessary to test calibration.

67. The “one error per line” problem

If an examination guarantees one error per line, use that rule. But do not generalize it to other tasks. Real language and many editing formats include correct lines.

68. The “no error” answer must feel legitimate

Train learners to submit “no change” confidently when analysis supports it.

Without this practice, they may feel compelled to manufacture a correction.

69. No-Change Drill A

All are correct. The entire drill tests restraint.

1. The series of workshops begins next month.
2. A number of volunteers are waiting outside.
3. The number of volunteers is increasing.
4. The equipment was expensive.
5. The news was surprising.
6. The fabric feels soft.
7. The speaker sounded confident.
8. She has lived here since 2021.
9. The train leaves at six tomorrow.
10. When the meeting begins, we will review the figures.
11. The results were stored securely.
12. The students discussed the issue.
13. Leonie is capable of finishing the task.
14. Iona looks forward to presenting her findings.
15. Everyone should bring their notes.
16. Who did you speak to?
17. The data are consistent with the model.
18. The team are arguing among themselves.
19. Only then did Maren understand the pattern.
20. Had the warning arrived earlier, the team might have changed course.

70. No-Change Drill B — context-dependent controls

These are correct in their intended contexts:

1. The report was complete yesterday.
2. The report was completed yesterday.
3. Before the bus arrived, we waited for an hour.
4. Before the bus arrived, we had been waiting for an hour.
5. The team is meeting now.
6. The team are arguing among themselves.
7. The data is useful.
8. The data are useful.
9. This result is different from the first.
10. This result is different to the first.

The learner must use target-standard and meaning evidence rather than assuming only one version can ever be correct.

71. The accepted-variation register

Maintain a small list of forms that vary across standard Englishes and target conventions:

• collective-noun agreement;
• data singular/plural;
• different from/to/than;
• toward/towards;
• at the weekend/on the weekend;
• have got/gotten distinctions;
• certain punctuation conventions.

Do not turn this list into exam-room debate. Resolve the target standard in advance.

72. The target-standard note

A learner’s editing card can include:

In this examination, follow the prescribed standard. Outside the exam, remember that other standard variants may exist.

73. False positives caused by memorized triggers

Common trigger errors:

• of appears → learner forces singular;
• before appears → learner forces past perfect;
• verb appears → learner forces adverb after it;
• passive appears → learner rewrites active;
• pronoun appears → learner searches for gender agreement even when singular they is standard.

Triggers should begin inspection, not determine the answer.

74. Trigger-Control Set

1. A number of students are waiting.
2. The quality of the images is high.
3. Before the teacher arrived, the students began.
4. The fabric feels soft.
5. The samples were stored safely.
6. Everyone should bring their notes.
7. The data are convincing.
8. Who did you speak to?
9. Only after the meeting did the reason become clear.
10. The team are disagreeing among themselves.

All are correct in their intended standards and contexts.

75. The evidence ladder

Rank evidence from strongest to weakest:

1. explicit task instruction;
2. clear grammatical relationship;
3. passage meaning and reference;
4. established lexical frame;
5. target-standard usage;
6. personal familiarity;
7. mere preference.

Do not allow preference to override grammar.

76. Evidence-Ladder Exercise

Sentence:

The team are discussing their individual concerns.

Possible evidence:

• British English can use plural collective agreement when members are foregrounded;
• the plural pronoun their supports member interpretation;
• the target exam may nevertheless prescribe singular-unit agreement.

Decision: follow the target standard, but do not call the plural form universally ungrammatical.

77. The repair threshold

Use a simple three-level threshold:

• Strong evidence: repair.
• Moderate evidence: inspect context or return.
• Weak evidence: preserve.

78. Strong-evidence examples

• may indicates → modal + base violation.
• has chose → perfect participle violation.
• the list…are → head agreement violation.
• interested on → established frame under target standard.

79. Moderate-evidence examples

• collective-noun agreement;
• data agreement;
• simple past versus past perfect where sequence is already explicit;
• compare to versus with;
• style-sensitive redundancy.

80. Weak-evidence examples

• “I would not write it this way.”
• “It sounds formal.”
• “I saw another version before.”
• “This word appears twice.”
• “Passive voice feels weak.”

81. Maren’s over-editing pattern

Maren is strong at sentence architecture. Her false-positive risk comes from wanting to improve structure beyond the task. She sees a grammatical sentence and imagines a more elegant one.

Her rule:

If the structure is grammatical and meaning is clear, style preference alone does not authorize a correction.

82. Iona’s over-editing pattern

Iona is strong at evidence and meaning. Her risk comes from exploring too many alternative analyses. She can talk herself out of a correct answer because she sees possible nuance.

Her rule:

Once the relevant relationship survives and no stronger contradiction appears, preserve the original.

83. Leonie’s over-editing pattern

Leonie is strong at execution. Her risk appears late: time remains, so she reopens answers and changes them without new evidence.

Her rule:

No new evidence, no new answer.

84. Personal false-positive map

Trigger	Wrong habit	Correct check
long subject	follow nearest noun	find head
before/after	force past perfect	ask whether sequence needs marking
linking verb	force adverb	ask whether complement describes subject
collective noun	force singular	check meaning + target standard
passive voice	rewrite active	ask whether grammar actually fails
sentence-final preposition	move preposition automatically	check register requirement

85. False-positive error log

For every unnecessary change, record:

• original correct form;
• change made;
• reason you thought it was wrong;
• actual rule;
• future preservation cue.

The later dedicated editing error-log spoke will systematize logging, but false-positive records are especially useful here.

86. Example false-positive log entry

Original: The fabric feels soft.
Changed to: softly.
Why: assumed verb needs adverb.
Actual rule: feels is linking verb; adjective describes fabric.
Future cue: linking verb → adjective possible.

87. Example variation log entry

Original: The data are convincing.
Changed to: is.
Why: learned mass-noun use.
Actual rule: plural scientific usage is standard.
Future cue: target standard before changing data.

88. Example tense false-positive log entry

Original: Before the teacher arrived, the students began the warm-up.
Changed to: had begun.
Why: assumed earlier past always requires past perfect.
Actual rule: simple past is possible because before already orders events.
Future cue: past perfect when relationship needs explicit marking, not automatically.

89. The preserve-or-change matrix

Original status	Your evidence	Action
Clearly wrong	Strong	Change
Probably wrong	Moderate	Check context / return
Looks unusual	Weak	Keep
Correct but stylistically different	None	Keep
Accepted variation	Target standard unresolved	Do not force change

90. Accuracy Practice Bank 1 — Agreement and number

Mark K or C.

1. The quality of the images is excellent.
2. The quality of the images are excellent.
3. A number of applicants have withdrawn.
4. A number of applicants has withdrawn.
5. The number of applicants has increased.
6. The number of applicants have increased.
7. Each of the machines needs repair.
8. Each of the machines need repair.
9. The news was surprising.
10. The news were surprising.
11. The equipment was stored safely.
12. The equipment were stored safely.
13. The students who work here have finished.
14. The students who works here have finished.
15. One of the students who work here has finished.
16. One of the students who works here have finished.
17. The series of lectures begins next week.
18. The series of lectures begin next week.
19. The effects of the change are visible.
20. The effects of the change is visible.

Key

Odd-numbered items K. Even-numbered items C in the intended standard, except 16 requires two relationships to be checked: relative-clause agreement and main agreement.

91. Accuracy Practice Bank 2 — Verb chains

1. The result may indicate a problem.
2. The result may indicates a problem.
3. The committee has chosen a chair.
4. The committee has chose a chair.
5. The students are revising.
6. The students are revise.
7. The samples were stored safely.
8. The samples were store safely.
9. They should have checked the answer.
10. They should have check the answer.
11. Leonie can finish on time.
12. Leonie can finishes on time.
13. The device is operating normally.
14. The device is operate normally.
15. The bridge was built in 1980.
16. The bridge was build in 1980.
17. She might have forgotten the key.
18. She might has forgotten the key.
19. The work has been completed.
20. The work has been complete.

Key

Odd K; even C, with item 20 potentially context-dependent if complete is interpreted adjectivally. This is precisely why meaning matters.

92. Accuracy Practice Bank 3 — Tense and aspect

1. She submitted the form yesterday.
2. She has submitted the form yesterday.
3. We have lived here since 2022.
4. We live here since 2022.
5. When the meeting begins, we will start.
6. When the meeting will begin, we will start.
7. By the time I arrived, they had left.
8. By the time I arrived, they have left.
9. Before the teacher arrived, the students began.
10. Before the teacher arrived, the students had begun.
11. The museum opened in 1998.
12. The museum has opened in 1998.
13. The museum has expanded several times since 1998.
14. The museum expanded several times since 1998 and continues to expand.
15. The train leaves at six tomorrow.
16. The train will leaves at six tomorrow.
17. Water freezes at 0°C.
18. Water is freezing at 0°C as a universal statement.
19. I am thinking about the problem.
20. I think you are right.

Key

1 K, 2 C, 3 K, 4 C, 5 K, 6 C, 7 K, 8 C, 9 K, 10 K, 11 K, 12 C, 13 K, 14 usually C in the stated ongoing meaning, 15 K, 16 C, 17 K, 18 C as a universal statement, 19 K, 20 K.

93. Accuracy Practice Bank 4 — Word form

1. The machine operates reliably.
2. The machine operates reliable.
3. The machine is reliable.
4. The machine is reliably.
5. The fabric feels soft.
6. The fabric feels softly.
7. The student answered confidently.
8. The student answered confident.
9. The speaker sounded confident.
10. The speaker sounded confidently.
11. The explanation was remarkably clear.
12. The explanation was remarkable clear.
13. The result changed suddenly.
14. The result changed sudden.
15. The students were bored.
16. The students were boring by the lecture.
17. The lesson was boring.
18. The lesson was bored.
19. The team reached a decision.
20. The team reached a decide.

Key

Odd K. Even C in the intended meanings.

94. Accuracy Practice Bank 5 — Prepositions and complements

1. Iona is interested in grammar.
2. Iona is interested on grammar.
3. Leonie is responsible for timing.
4. Leonie is responsible of timing.
5. The result depends on several factors.
6. The result depends of several factors.
7. The students listened to the recording.
8. The students listened the recording.
9. Maren is capable of explaining the answer.
10. Maren is capable to explain the answer.
11. Iona looks forward to presenting her findings.
12. Iona looks forward to present her findings.
13. Leonie avoided repeating the mistake.
14. Leonie avoided to repeat the mistake.
15. Maren decided to revise the paragraph.
16. Maren decided revising the paragraph.
17. The tutor made the students explain their reasoning.
18. The tutor made the students to explain their reasoning.
19. The school allowed students to use dictionaries.
20. The school allowed students use dictionaries.

Key

Odd K; even C.

95. Accuracy Practice Bank 6 — Relative clauses and pronouns

1. The students who study regularly improve.
2. The students who studies regularly improve.
3. The student who studies regularly improves.
4. The student who study regularly improves.
5. The researcher who designed the study presented the results.
6. The researcher who she designed the study presented the results.
7. Between you and me, the plan seems risky.
8. Between you and I, the plan seems risky.
9. Please contact me if you have questions.
10. Please contact myself if you have questions.
11. Everyone should bring their notes.
12. Everyone should bring his or her notes.
13. The teacher spoke to Maren and him.
14. The teacher spoke to Maren and he.
15. One of the researchers who work here has published a paper.
16. One of the researchers who works here have published a paper.
17. The books that are on the shelf belong to Iona.
18. The books that is on the shelf belong to Iona.
19. The teacher whom we met was helpful.
20. The teacher who we met was helpful.

Key

1 K, 2 C, 3 K, 4 C, 5 K, 6 C, 7 K, 8 C, 9 K, 10 C, 11 K in modern standard English, 12 K though stylistically heavier, 13 K, 14 C, 15 K in the intended plural-antecedent reading, 16 C, 17 K, 18 C, 19 K, 20 may also be acceptable in many modern contexts where who is used for object position informally and increasingly in standard usage; follow target standard.

96. Accuracy Practice Bank 7 — Modifiers and comparison

1. Walking through the museum, Maren noticed the paintings.
2. Walking through the museum, the paintings impressed Maren.
3. After reading the report, Iona understood the problem.
4. After reading the report, the problem became obvious to Iona.
5. This method is easier.
6. This method is more easier.
7. No other runner was as fast as Leonie.
8. No other runner was as faster as Leonie.
9. The salary of engineers is higher than that of teachers.
10. The salary of engineers is higher than teachers.
11. The rainfall here is greater than that on the coast.
12. The rainfall here is greater than the coast.
13. Only Maren revised the answer.
14. Maren only revised the answer.
15. Maren revised only the answer.
16. The second passage is the shortest.
17. The second passage is the most shortest.
18. Exhausted by the climb, the hikers rested.
19. Built in 1920, the bridge still attracts tourists.
20. Built in 1920, tourists still visit the bridge.

Key

1 K, 2 C, 3 K, 4 C if the introductory phrase is intended to modify Iona, 5 K, 6 C, 7 K, 8 C, 9 K, 10 C, 11 K, 12 C, 13–15 can all be grammatical but express different scopes, 16 K, 17 C, 18 K, 19 K, 20 C because the modifier illogically attaches to tourists.

97. Accuracy Practice Bank 8 — Parallelism, connectors and boundaries

1. The programme aims to reduce waste, improve efficiency and lower costs.
2. The programme aims to reduce waste, improving efficiency and lower costs.
3. Maren enjoys reading, writing and discussing ideas.
4. Maren enjoys reading, writing and to discuss ideas.
5. Although the road was closed, we continued.
6. Although the road was closed, but we continued.
7. Because the printer failed, the report was delayed.
8. Because the printer failed, therefore the report was delayed.
9. The evidence was limited; however, the conclusion was cautious.
10. The evidence was limited, however the conclusion was cautious.
11. The alarm sounded; everyone left.
12. The alarm sounded, everyone left.
13. Because the weather changed suddenly, we turned back.
14. Because the weather changed suddenly.
15. The team collected samples, analysed them and wrote the report.
16. The team collected samples, analysed them and writing the report.
17. The task was difficult, but Leonie remained calm.
18. The task was difficult, therefore Leonie remained calm.
19. The first trial failed, so the team repeated it.
20. The first trial failed because the team repeated it.

Key

1 K, 2 C, 3 K, 4 C, 5 K, 6 C, 7 K, 8 C, 9 K, 10 usually C in formal punctuation if however is a conjunctive adverb joining two independent clauses, 11 K, 12 C, 13 K, 14 C fragment, 15 K, 16 C, 17 K, 18 C because logical relation is wrong, 19 K, 20 C because cause/effect reverses.

98. Accuracy Practice Bank 9 — Accepted variation

Do not force a change without target-standard evidence.

1. The data are convincing.
2. The data is convincing.
3. The team is meeting today.
4. The team are arguing among themselves.
5. This is different from the first version.
6. This is different to the first version.
7. This is different than the first version.
8. Who did you speak to?
9. To whom did you speak?
10. Everyone should bring their notes.
11. Everyone should bring his or her notes.
12. The report was carefully analysed.
13. The researchers carefully analysed the report.
14. Before the meeting began, we reviewed the notes.
15. Before the meeting had begun, we reviewed the notes.
16. The results were compared with earlier findings.
17. The results were compared to earlier findings.
18. The staff are reviewing their individual notes.
19. The staff is reviewing the schedule as one unit.
20. The programme will start on Monday.

Several of these vary by dialect, meaning or style but remain grammatical.

99. Accuracy Practice Bank 10 — Style traps

Assume a grammar-only task. Which need change?

1. The report was written by the team.
2. The team wrote the report.
3. The explanation was extremely clear.
4. The explanation was very clear.
5. The students completed the task successfully.
6. The students successfully completed the task.
7. The sentence is long but grammatical.
8. The sentence is short and direct.
9. The results were carefully analysed.
10. The researchers analysed the results carefully.
11. The list of names are complete.
12. The equipment were expensive.
13. The students discussed about the issue.
14. Who did you speak to?
15. Everyone should bring their notes.
16. The data are convincing.
17. The paragraph repeats one technical term three times.
18. The passive voice is used twice.
19. The writer chose a formal register.
20. The writer chose an informal but grammatical contraction.

Key

11, 12 and 13 clearly require grammatical repair in the intended standard. The others may be stylistic choices or accepted variation unless the task scope says otherwise.

100. The personal false-positive profile

After these banks, classify unnecessary changes by trigger:

• agreement trigger;
• tense trigger;
• linking-verb trigger;
• passive trigger;
• variation trigger;
• formal-structure trigger;
• preposition preference;
• style preference;
• late answer change.

101. High-confidence false positives

These are especially important because the learner will repeat them unless the misconception is corrected.

Example:

The fabric feels soft.

If the learner confidently changes soft → softly, teach the linking-verb relationship directly.

102. Low-confidence false positives

These often arise from uncertainty rather than misconception.

Example:

The data are convincing.

The learner hesitates, changes to is, then changes back. Training should build target-standard knowledge and stop rules.

103. Late false positives

Track whether unnecessary changes happen mostly during review.

If so, the problem may be answer-changing discipline rather than grammar knowledge.

104. Review-value audit

For five timed tasks, record:

• answers changed during review;
• wrong → right changes;
• right → wrong changes;
• wrong → different wrong changes;
• time spent reviewing.

This reveals whether review adds value.

105. Productive review

Productive review finds new evidence:

• missed subject head;
• missed time marker;
• wrong answer transfer;
• forgotten lexical frame;
• clear reference issue.

106. Unproductive review

Unproductive review:

• rereads secure items repeatedly;
• changes answers to feel active;
• rewrites style;
• reopens accepted variation;
• chases unfamiliar but grammatical forms.

107. The review hierarchy

When time remains:

1. check unanswered items;
2. check marked uncertainty;
3. check answer transfer;
4. verify high-value repairs;
5. stop.

Do not re-solve every item equally.

108. The final-minute rule

In the final minute, do not introduce a large speculative rewrite unless the original error is unmistakable.

109. The stop point

Stop editing when:

• the relevant relationships have been checked;
• uncertain items received one purposeful return;
• answer transfer is complete;
• remaining concerns are stylistic or unsupported;
• continued checking is mostly repetition.

110. Why stopping can improve accuracy

Every additional inspection creates opportunity for both discovery and damage. Once expected benefit falls below expected risk, stopping is rational.

111. Accuracy Passage A — museum archive

A museum archive contains thousands of photographs. Each image is labelled with a date, location and reference number. Since the archive introduced a digital catalogue, researchers have found material more quickly. A number of volunteers are helping with scanning. The number of fragile photographs are still increasing as new donations arrive. Staff are responsible for checking each item before it is handled.

Key

Only The number…are → is clearly requires repair in the intended standard. Preserve the rest.

112. Accuracy Passage B — weather monitoring

A research team has monitored the site since 2021. Every week, technicians visit three stations and download readings. The equipment was upgraded last year, and the new sensors have improved data quality. The team are currently comparing their individual field notes. The results from one station appears unusual and will be checked again.

Key

The clear error is results…appears → appear. team are may be legitimate member-focused plural collective agreement depending on target standard.

113. Accuracy Passage C — theatre rehearsal

An orchestra rehearses several times before a public performance. The conductor may stop to repeat a difficult transition, and musicians often mark changes in their scores. Before the final rehearsal began, the players reviewed the most difficult passages. The orchestra has improved steadily since Monday. The musicians are now familiar with the transitions that caused problems earlier.

No changes are required.

114. Accuracy Passage D — digital mapping

Researchers use digital maps to combine photographs, coordinates and field notes. Each record contains information from one site. Since the project started, teams collected more than ten thousand observations. The data are stored on secure servers and can be searched by date or location. One of the most useful features are the ability to compare old and new images.

Key

Since…teams collected → have collected if the project continues to the present. One…features are → is. data are is correct in plural scientific usage.

115. Accuracy Passage E — reading group

A reading group meets every Thursday. Each member brings one question to the discussion. Since Iona joined, she has become more confident explaining difficult passages. Maren is especially good at identifying how sentences are structured, while Leonie focuses on timing. Everyone should bring their notes so that references can be checked quickly.

No changes are required.

116. Accuracy Passage F — engineering test

Engineers test materials to understand how they behave under repeated load. A sample is placed in a machine that applies controlled force. Sensors records how far the material bends. Since the laboratory installed new equipment, researchers have measured smaller changes. The results are useful only if the procedure is followed consistently.

Key

Only Sensors records → Sensors record. Preserve everything else.

117. Accuracy Passage G — history project

A group of students are creating a digital history of their neighbourhood. They have interviewed residents, scanned photographs and visited archives. Each interview provides details that may not appear in official records. Since the project began, the group has learned to distinguish memory from evidence. Their teacher is pleased with the progress.

Key

If plural collective agreement is acceptable, A group of students are creating… can be correct. Do not force singular unless target standard requires it. Other sentences are correct.

118. Accuracy Passage H — food testing

Food scientists compare products stored under different conditions. Each sample is weighed before and after storage. Since the laboratory introduced digital scales, measurements have become more precise. The team is now capable to detect smaller changes than before. The value of the experiment depends on using identical procedures for every sample.

Key

Only capable to detect → capable of detecting.

119. Accuracy Passage I — transport study

A transport team uses passenger data to understand where crowding occurs. The data is collected from ticket gates, vehicle counters and surveys. Each source provides a different view of the network. Since the team introduced a new dashboard, planners have identified patterns more quickly. However, the team remains aware of the limits of the data.

No change is required in a general-English standard that treats data as a singular mass noun. A scientific or institution-specific style guide may prefer plural.

120. Accuracy Passage J — conservation trail

A conservation trail passes through several habitats. Visitors are asked to remain on marked paths because repeated foot traffic can damage fragile areas. Since new signs were installed, fewer people have left the main trail. Rangers are pleased about the improvement but still monitor several sections closely. The condition of the trail surfaces varies after heavy rain.

No clear grammatical changes are required. Pleased about is possible, though pleased with may be preferred in some contexts.

121. Accuracy Passage K — astronomy night

A school astronomy night gives students an opportunity to observe objects they usually see only in diagrams. Each telescope is aimed at a different target, and volunteers explain what visitors are seeing. If clouds move across the sky, the programme may shift indoors. Since the event began five years ago, attendance has increased steadily. Organisers are pleased by the interest.

No change is required.

122. Accuracy Passage L — recipe testing

Recipe testing involves more than checking whether food tastes good. A reliable recipe needs clear quantities, timings and instructions. During testing, each version is prepared under similar conditions. If a cake rises unevenly, testers may adjust temperature, mixing time or ingredient ratios. Since one bakery introduced a formal testing process, it has reduced waste and improved consistency.

No change is required.

123. Accuracy Passage M — travel writing

Good travel writing depends on observation rather than a list of attractions. A writer may notice how a street sounds in the morning, how people use a square or how weather changes a place. Since Leonie began keeping shorter field notes, her descriptions have become more specific. She is also more aware of which details support the main idea.

No change is required.

124. Accuracy Passage N — school timetable

A school timetable looks simple when printed, but creating one requires many constraints to be balanced. Each class needs rooms, teachers and enough time for different subjects. Some lessons require specialist spaces that are shared by several groups. Since the school added new courses this year, planners have adjusted the schedule several times. Small changes can have surprisingly wide effects.

No change is required.

125. Accuracy Passage O — air-quality sensors

Low-cost air-quality sensors have made local monitoring more accessible. A single sensor can collect large amounts of data, but the readings are not always accurate without calibration. Researchers are interested in comparing low-cost devices with reference instruments. Since one city installed a network last year, officials have used the data to identify several hotspots.

No change is required.

126. Accuracy Passage P — archive control

A theatre company has kept programmes, posters and costume sketches for more than fifty years. The collection was stored in several rooms until a recent project brought the material together. Volunteers have begun creating a catalogue that describes each item. Before an object is entered in the catalogue, its condition is recorded in detail. The team hopes the system will make the archive easier to use.

No change is required.

127. Accuracy Passage Q — controlled error

A theatre company has kept programmes, posters and costume sketches for more than fifty years. The collection were stored in several rooms until a recent project brought the material together. Volunteers have begun creating a catalogue that describes each item. Before an object is entered in the catalogue, its condition is recorded in detail. The team hopes the system will make the archive easier to use.

Key

Only collection were → collection was. This passage is designed to test restraint around one obvious error.

128. Accuracy Passage R — accepted variation control

The research team is reviewing its results this week. Staff are checking the raw data while the project manager prepares a summary. The data are stored in two secure locations. A number of analysts have also repeated the main calculations. The committee will meet on Friday.

This passage can be fully correct in a standard that permits plural staff and plural scientific data.

129. Accuracy Passage S — correct but formal

Only after the second trial did the researchers identify the source of the error. Had they stopped after the first trial, the problem might have remained hidden. Not only did the revised method improve accuracy, but it also reduced testing time.

No change is required.

130. Accuracy Passage T — one hidden error

Only after the second trial did the researchers identify the source of the error. Had they stopped after the first trial, the problem might remained hidden. Not only did the revised method improve accuracy, but it also reduced testing time.

Key

might remained → might have remained. Preserve the surrounding formal structures.

131. Why near-identical passages are useful

Pairs such as Q/P and S/T force the learner to distinguish the actual error from the surrounding language. The learner cannot rely on “this sentence type usually has a problem.”

132. Minimal-Pair Accuracy Set A

1. The list of changes is complete.
2. The list of changes are complete.
3. The fabric feels soft.
4. The fabric feels softly.
5. The result may indicate a problem.
6. The result may indicates a problem.
7. The students who study regularly improve.
8. The students who studies regularly improve.
9. The report was based on evidence.
10. The report was based in evidence.

For each pair, the first is correct and the second requires repair.

133. Minimal-Pair Accuracy Set B

1. Before the meeting began, we reviewed the notes.
2. Before the meeting had begun, we reviewed the notes.
3. The data are convincing.
4. The data is convincing.
5. The team is meeting today.
6. The team are arguing among themselves.
7. Who did you speak to?
8. To whom did you speak?
9. Everyone should bring their notes.
10. Everyone should bring his or her notes.

All can be correct in appropriate meanings and standards. This set trains tolerance of variation.

134. The “correct but not my choice” drill

Read each sentence and say:

I might write it differently, but the original is grammatical, so I will leave it.

This is especially useful for advanced students.

135. Grammar versus style table

Sentence feature	Grammar issue?	Style issue?
passive voice	not automatically	sometimes
long sentence	not automatically	sometimes
repeated technical noun	not automatically	context-dependent
agreement mismatch	yes	also affects style
wrong complement	yes	yes
sentence-final preposition	not automatically	register-dependent
split infinitive	not automatically	preference-dependent

136. Calibration drill: justify “keep”

For each correct sentence, explain why it should stay:

1. The fabric feels soft. — linking verb + adjective.
2. A number of students are waiting. — idiomatic quantity phrase with plural agreement.
3. The data are convincing. — standard plural scientific usage.
4. Who did you speak to? — standard sentence-final preposition.
5. Before the teacher arrived, the students began. — simple past sequence already clear.

137. Calibration drill: justify “change”

1. The list of names are complete. — singular head list.
2. The machine operates reliable. — adverb slot.
3. She has submitted yesterday. — finished past time.
4. Interested on grammar. — adjective frame.
5. The students who studies… — plural antecedent controls relative-clause verb.

138. The two-proof rule

For high-risk false-positive learners, require two proofs before changing a low-confidence item:

1. name the relationship;
2. show the contradiction.

Example:

The list of names are complete.

Relationship: subject–verb agreement.
Contradiction: head list singular, verb are plural.

Now change.

139. The one-proof rule for high-confidence errors

Obvious errors do not need over-analysis. If the relationship is clear, one proof is enough.

may indicates → modal + base → indicate.

140. Do not make restraint slow

False-positive control should not turn into endless caution. The goal is fast evidence-based preservation.

141. Fast preservation cues

• “No failed relationship.”
• “Variant accepted.”
• “Style only.”
• “Meaning changes if edited.”
• “Trigger but no contradiction.”

142. Accuracy under time

In timed work, preserve a correct line once the relevant checks survive. Do not spend equal time proving every correct sentence exhaustively.

143. The high-risk review list

During final review, revisit only:

• marked uncertainty;
• answers with low confidence;
• known personal false-positive families;
• answer-transfer risks.

144. The low-risk review list

Do not reopen:

• obvious modal repairs;
• secure agreement decisions;
• clear lexical frames;
• already verified answer transfers.

145. The accuracy scorecard

Metric	Meaning
Detection rate	genuine errors found
Repair rate	found errors fixed correctly
False-positive rate	correct forms wrongly changed
Precision	share of edits that were justified
Preservation score	correct forms left intact
Right→wrong review changes	damage caused by second-guessing
Time sinks	items consuming disproportionate time

146. Example scorecard A

Errors present: 10. Errors found: 9. Repairs correct: 8. Correct items presented: 20. Correct items changed: 5.

• Detection: 90%.
• Repair among detected: 8/9.
• False-positive rate: 25%.

Diagnosis: strong detection, weak restraint.

147. Example scorecard B

Errors present: 10. Errors found: 6. Repairs correct: 6. Correct items presented: 20. Correct items changed: 0.

Diagnosis: excellent precision, weak recall. Training should increase detection without sacrificing restraint.

148. Example scorecard C

Errors present: 10. Errors found: 9. Repairs correct: 9. Correct items changed: 0.

This is strong editing accuracy. Next test: unseen passage, realistic time, accepted variation.

149. Accuracy versus score stability

One accurate set is not enough. Precision must survive across different topics and days. The later lane spoke on score stability will own variance in depth.

150. One-week false-positive repair cycle

Day 1 — baseline

Mixed correct/error set. Record false-positive triggers.

Day 2 — correct-control families

Use sentences containing the learner’s top three triggers but no errors.

Day 3 — minimal pairs

Correct and incorrect versions side by side.

Day 4 — accepted variation

Practise target-standard versus globally accepted forms.

Day 5 — low-density passage

Only one or two real errors.

Day 6 — timed mixed set

Apply new-evidence answer-changing rule.

Day 7 — delayed unseen retest

Check whether unnecessary edits decrease.

151. Six-week progression

Week 1: measure false positives. Week 2: controller and evidence thresholds. Week 3: correct-control grammar families. Week 4: accepted variation and style/grammar separation. Week 5: low-density passages and review discipline. Week 6: timed unseen accuracy and delayed retest.

152. Twelve-week progression

Weeks 1–2 establish precision metrics. Weeks 3–4 reduce high-confidence misconceptions. Weeks 5–6 train low-confidence correct forms. Weeks 7–8 add accepted variation and advanced syntax. Weeks 9–10 use low-density passage editing. Week 11 uses authentic simulation. Week 12 compresses the learner’s final preservation rules.

153. A 30-minute session

1. 5 min: retrieve three personal false-positive traps;
2. 8 min: correct-control set;
3. 5 min: mixed minimal pairs;
4. 5 min: one low-density passage;
5. 5 min: review right→wrong changes;
6. 2 min: update preservation cue.

154. A 60-minute session

1. 10 min: evidence-threshold review;
2. 15 min: mixed correct/error sentences;
3. 10 min: accepted variation;
4. 10 min: low-density passage;
5. 10 min: timed review and answer-changing audit;
6. 5 min: update false-positive map.

155. Teacher diagnostic questions

• Does the learner change correct forms?
• Which correct structures trigger those changes?
• Do false positives happen mainly during first pass or review?
• Does time pressure increase them?
• Are they high-confidence misconceptions or low-confidence guesses?
• Does the learner distinguish grammar from style?
• Does the learner know target-standard variation?

156. Parent-facing explanation

A useful explanation is:

Your child is not losing editing marks only because errors are missed. Some marks are being lost because correct sentences are changed. We are training two skills at the same time: finding real errors and learning when to leave correct English alone.

157. Student-facing explanation

Your job is not to prove that you can rewrite English. Your job is to prove that you can tell the difference between a real grammatical failure and a sentence that is simply unfamiliar.

158. Accuracy Mastery Passage 1

A school library has expanded its digital services over the past three years. The number of students using the online catalogue has increased steadily. A number of teachers are also using digital collections in class. Each of the new resources includes a short guide. The library team are currently discussing their individual feedback before the next update.

No clear change is required if plural member-focused collective agreement is allowed for team are. Otherwise follow the target standard. All other structures are correct.

159. Accuracy Mastery Passage 2

A weather station records temperature, wind and pressure throughout the day. Each sensor is protected by a small housing. Since the station was upgraded, data transmission has become more reliable. The data are stored automatically and checked when unusual values appear. Technicians are responsible for visiting the site when remote checks cannot resolve a problem.

No change is required.

160. Accuracy Mastery Passage 3

A theatre company has kept programmes and photographs for more than fifty years. The collection was stored in several rooms until a recent project brought the material together. Volunteers have begun creating a catalogue that describes each item. Before an object is entered in the catalogue, its condition are recorded in detail. Researchers can now search the archive by date and production.

Key

Only condition are → condition is.

161. Accuracy Mastery Passage 4

A school debate team meets twice a week. Each speaker receives a fixed amount of preparation time. Since Leonie began timing her practice, she has become more consistent at finishing her conclusions. Maren is responsible for checking argument structure, while Iona listens for unsupported claims. When the next competition begins, the team will use the same routine.

No change is required.

162. Accuracy Mastery Passage 5

A digital archive stores photographs of historic buildings. Each building is photographed from the same viewpoints so changes can be compared. Since the project started, volunteers have documented more than two hundred sites. The images are stored with notes about date and location. Researchers are interested on how neighbourhoods change over long periods.

Key

Only interested on → interested in.

163. Accuracy Mastery Passage 6

A museum education team designs activities that help visitors look closely at objects. One activity asks participants to spend two minutes describing what they see before reading the label. Since the activity was introduced, educators have used it with school groups, families and adults. The team is interested in whether the method changes how long visitors spend with an object. Early feedback has been encouraging.

No change is required.

164. Accuracy Mastery Passage 7

Engineers monitoring a tunnel use sensors to track movement in the surrounding ground. A network of instruments is installed at different depths. Each device sends readings to a central system. Since construction began, the network has provided continuous data. However, one unusual value do not automatically indicate a serious problem.

Key

Only one unusual value do → does.

165. Accuracy Mastery Passage 8

A radio producer prepares a programme by combining interviews, narration and recorded sound. Every segment needs a clear purpose so that the final programme does not feel crowded. Producers often record more material than they will use. Since digital editing became easier, teams have experimented with more complex sound design. However, complexity is not automatically effective.

No change is required.

166. Accuracy Mastery Passage 9

A research greenhouse allows scientists to control light, water and temperature. Each growing area has separate sensors. Since the greenhouse was renovated, researchers have used more precise lighting schedules. Plants are measured regularly, and images are taken from the same angle each week. The consistency of the method help researchers compare growth over time.

Key

Only method help → helps.

167. Accuracy Mastery Passage 10

A language exchange programme pairs students who speak different first languages. Each participant meets a partner once a week. The aim is not to teach formal lessons but to practise conversation and learn about everyday expressions. Since the programme started, participants have reported greater confidence in speaking. Organisers are careful to match partners with compatible schedules.

No change is required.

168. Mastery criterion

False-positive control is stable when the learner can:

• maintain high error detection;
• leave correct lines intact;
• explain accepted variation;
• separate grammar from style;
• change answers only with new evidence;
• preserve correct advanced structures;
• perform similarly under time.

169. Independence test

The learner is independent when she can say:

I changed this because the subject head is singular and the verb is plural. I left that one because the sentence is grammatical even though I would phrase it differently. I left this collective noun because the target standard accepts the form. I changed my earlier answer only because I found new evidence.

170. Final 50-item accuracy audit

Mark K or C. Do not assume half must be wrong.

1. The list of names is complete.
2. The list of names are complete.
3. A number of students are waiting.
4. The number of students is increasing.
5. The result may indicate a problem.
6. The result may indicates a problem.
7. The committee has chosen a chair.
8. The committee has chose a chair.
9. The samples were stored safely.
10. The samples were store safely.
11. She submitted the report yesterday.
12. She has submitted the report yesterday.
13. We have lived here since 2022.
14. We live here since 2022.
15. When the meeting starts, we will begin.
16. When the meeting will start, we will begin.
17. Before the teacher arrived, the students began.
18. Before the teacher arrived, the students had begun.
19. The fabric feels soft.
20. The fabric feels softly.
21. The machine operates reliably.
22. The machine operates reliable.
23. Iona is interested in grammar.
24. Iona is interested on grammar.
25. Leonie is capable of explaining the answer.
26. Leonie is capable to explain the answer.
27. The students who study regularly improve.
28. The students who studies regularly improve.
29. Everyone should bring their notes.
30. Who did you speak to?
31. Only after the second test did the fault become clear.
32. Only after the second test the fault became clear.
33. The data are convincing.
34. The data is convincing.
35. The team is meeting today.
36. The team are arguing among themselves.
37. The programme aims to reduce waste, improve efficiency and lower costs.
38. The programme aims to reduce waste, improving efficiency and lower costs.
39. The rainfall here is higher than that on the coast.
40. The rainfall here is higher than the coast.
41. Walking through the museum, Maren noticed the paintings.
42. Walking through the museum, the paintings impressed Maren.
43. Although the road was closed, we continued.
44. Although the road was closed, but we continued.
45. The alarm sounded; everyone left.
46. The alarm sounded, everyone left.
47. The results were carefully analysed.
48. The researchers carefully analysed the results.
49. The report was written by the team.
50. The team wrote the report.

Audit key

1 K. 2 C. 3 K. 4 K. 5 K. 6 C. 7 K. 8 C. 9 K. 10 C. 11 K. 12 C. 13 K. 14 C. 15 K. 16 C. 17 K. 18 K. 19 K. 20 C in the intended linking meaning. 21 K. 22 C. 23 K. 24 C. 25 K. 26 C. 27 K. 28 C. 29 K. 30 K. 31 K. 32 C. 33 K in plural scientific usage. 34 K in common singular-mass usage. 35 K. 36 K in a variety permitting plural member-focused collective agreement. 37 K. 38 C. 39 K. 40 C. 41 K. 42 C. 43 K. 44 C. 45 K. 46 C. 47 K. 48 K. 49 K. 50 K.

171. Final accuracy scorecard

After the audit, calculate:

• genuine errors present;
• genuine errors detected;
• correct repairs;
• correct items changed;
• accepted variants changed;
• right→wrong answer changes;
• low-confidence correct answers;
• high-confidence wrong answers.

172. What improvement looks like

Improvement means:

• false-positive rate falls;
• precision rises;
• correct-line preservation rises;
• detection stays high;
• accepted variation no longer triggers unnecessary changes;
• review produces more wrong→right than right→wrong changes;
• time sinks decrease;
• the learner stops earlier when no new evidence exists.

173. What overcorrection looks like

Warning signs:

• more edits than genuine errors;
• frequent style rewrites;
• multiple changes to the same sentence;
• correct answers repeatedly changed during review;
• accepted variation treated as failure;
• long explanations without a specific grammatical contradiction.

174. The final personal preservation card

1. FIND A REAL RELATIONSHIP.
2. REQUIRE EVIDENCE.
3. KEEP CORRECT VARIANTS.
4. CHANGE AS LITTLE AS POSSIBLE.
5. VERIFY ONCE.
6. CHANGE AN ANSWER ONLY WITH NEW EVIDENCE.
7. STOP WHEN THE LANGUAGE SURVIVES.

175. Relation to the wider SETC system

This article specializes false-positive control in supplied-text editing. It does not replace the broader English Exam Calibration owner, which applies trust/check/change decisions across examination tasks. It also remains distinct from Proofreading Under Exam Pressure, which deals with checking the learner’s own writing. Here, the core risk is damaging a supplied text that was already correct.

176. Next step in the lane

Return to the Secondary English Editing Examination Performance hub. The next spoke is SETC-SEC-EDIT-080 — Proofreading vs Editing for English Exams | Know Which Errors the Task Actually Tests. That page will define the boundary between supplied-text editing, proofreading, self-editing, surface correction and broader revision.

177. Final principle

THE BEST EDITOR IS NOT THE PERSON WHO CHANGES THE MOST. THE BEST EDITOR CHANGES WHAT FAILS, PRESERVES WHAT WORKS, AND STOPS WHEN THE EVIDENCE RUNS OUT.