Secondary English Editing Examination Performance | Find, Prove and Repair Grammar Errors Under Test Conditions

Editing in an English examination is not a hunt for words that look strange. It is a controlled decision system: find a possible fault, prove which grammatical relationship has failed, repair the smallest unit that restores correctness, then verify that the sentence still means what the passage needs it to mean.

This hub is the canonical SETC route for Secondary English supplied-text editing examination performance. It is designed for learners anywhere in the world who are asked to identify grammatical errors in sentences or passages, decide whether a line is already correct, supply a correction, and do all of that under a finite clock. The exact layout of an examination may vary. Some papers use one error per line. Some use complete sentences. Some mix correct and incorrect items. Some ask for an incorrect word and its replacement; others ask for a rewritten sentence. The surface changes, but the underlying performance problem is remarkably stable.

The goal here is not to replace grammar teaching. Grammar knowledge matters. Nor is this the same as revising your own composition after writing it. SETC already has a broad guide to English Editing Performance for the final repair of your own completed examination writing. This lane owns a narrower and technically different job: editing a supplied text whose intended meaning you did not create. That changes the burden of proof. You are not free to rewrite for taste. You must infer what the writer is trying to say, identify the grammar failure, make the smallest justified correction, and avoid damaging language that was already correct.

The entire operating model can be compressed into one line:

READ FOR MEANING → LOCATE THE GRAMMATICAL RELATIONSHIP → TEST THE SUSPECT FORM → REPAIR MINIMALLY → REREAD IN CONTEXT → ACCEPT OR REJECT THE EDIT → MOVE ON.

Maren, Iona and Leonie will accompany the system. Maren naturally notices sentence structure and how parts fit together. Iona is strongest when a decision needs evidence: what does this pronoun refer to, what time relationship is implied, which noun controls this verb, what exactly is the sentence claiming? Leonie is interested in execution: how to scan, when to stop, how to recover after uncertainty, and how to keep accuracy alive when the clock is moving. The three lenses are useful because editing failure rarely comes from one cause. A learner can know a grammar rule yet fail to notice the error. Another can detect that something is wrong yet choose the wrong correction. A third can solve items correctly in isolation but lose accuracy in a passage under time.

What this lane owns

This lane owns the performance system around supplied-text English editing for secondary learners: error detection, grammatical diagnosis, correction, verification, false-positive control, passage reading, timed scanning, unseen transfer, personal error maps, practice design, mock editing papers, recovery, score stability, and the feedback loop that decides what to train next.

It does not try to absorb every neighbouring skill. For the wider examination system, return to English Examination Performance. For accuracy across grammar and vocabulary under pressure, use English Language Control. For turning lost marks into causes, use English Exam Error Analysis. For how drills should mature into authentic simulation, use English Exam Practice Design. For full-paper stress testing, use English Mock Exam Performance. For converting corrections into future improvement, use English Exam Feedback Transfer.

This hub specializes those larger systems for one deceptively small examination act: seeing exactly what is wrong in somebody else’s sentence without changing what is right.

The first distinction: editing is not proofreading, rewriting or taste

Students often use the words editing and proofreading as if they were interchangeable. In everyday publishing, the boundaries overlap. In examination tasks, the distinction matters because the marking target is often narrower than ordinary real-world editing.

Proofreading usually emphasizes surface correctness: spelling, punctuation, capitalization, typographical slips and obvious grammatical errors. Editing can be broader: grammar, sentence structure, word form, reference, cohesion, consistency and sometimes meaning. Rewriting gives the writer more freedom to improve expression. A supplied-text grammar editing task may permit none of that stylistic freedom. The examiner is often testing whether you can locate a specific grammatical fault and supply a constrained correction.

That means the question is not, “Can I make this sentence sound better?” It is, “Can I prove that a particular form is grammatically incompatible with the relationship the sentence expresses?”

Consider: The list of proposals were circulated yesterday. A learner who edits by sound may hesitate because the plural noun proposals sits next to were. A grammatical editor identifies the head of the subject noun phrase: list. The prepositional phrase of proposals modifies it. The finite verb must agree with list, so were becomes was. The correction is not a preference. It follows from the structure.

Now consider: The committee discussed the proposal yesterday. A learner who has been trained to expect one error per line may invent a change because the sentence looks too easy. Perhaps discussed about seems more familiar. That would create an error. Strong editing includes the ability to leave correct language alone.

The editing-performance chain

Strong performance can be modelled as seven stages. Each stage can fail independently.

1. Meaning acquisition. Understand what the sentence or passage is trying to say.
2. Error detection. Notice that a form or relationship deserves inspection.
3. Diagnosis. Name the grammatical relationship that is failing.
4. Candidate repair. Generate the smallest plausible correction.
5. Verification. Test grammar, meaning and local context after the repair.
6. Decision. Commit to the edit or retain the original if the evidence is insufficient.
7. Execution. Record the answer accurately and protect time for the rest of the task.

A learner who sees the wrong verb but cannot explain why it is wrong has a diagnosis weakness. A learner who can explain the rule after a teacher points at the line but fails to notice the line independently has a detection weakness. A learner who always finds errors but also changes correct lines has a false-positive problem. A learner who is accurate untimed but collapses in a mixed passage has a transfer or execution problem.

This distinction matters because “do more editing exercises” is not a diagnosis. Practice must target the failing stage.

Detection before repair

The first child article in this lane is English Error Correction Exercises With Answers | Train Detection Before Repair. That sequence is deliberate. Many worksheets present a sentence that is already labelled incorrect. The learner’s job is merely to fix it. Real examination editing is harder because the learner may first need to decide whether anything is wrong at all.

Detection is a search problem. You are scanning language for relationships with a known failure probability: subject and verb, time and tense, noun and determiner, preposition and complement, pronoun and antecedent, adjective and noun, adverb and verb, modal and base verb, auxiliary and participle, comparison and parallel term, conjunction and logical relation. Expert editors do not inspect every word with equal intensity. They recognize structures that deserve higher attention.

Maren calls this reading the sentence as a machine with joints. Iona asks what evidence controls each joint. Leonie asks how many such checks can be performed before the search becomes too slow.

Meaning first: grammar is not a bag of isolated rules

A supplied-text editor has a problem that a grammar drill often removes: context. The same word can be correct in one sentence and wrong in another. The same tense can be correct or incorrect depending on the timeline. The same pronoun can be clear in one passage and ambiguous in the next. Editing begins by reading for meaning because grammar encodes relationships in meaning.

Take: By the time the rescue team arrived, the hikers left the valley. The sentence may be grammatical in a rare interpretation, but the likely intended timeline is that the hikers departed before the rescue team arrived. The earlier action should therefore be marked as anterior to the later past event: had left. You need the timeline, not just a memorized instruction to “check tense.”

Or take: The scientist placed the samples in the freezer because they were unstable. The pronoun they plausibly refers to samples. But if the preceding sentence also mentioned chemicals and the next sentence clarifies that the chemicals were unstable, the reference might become ambiguous. Editing cannot be separated from local discourse.

The smallest meaning-preserving repair

A central rule of this lane is the minimal-repair principle: change the smallest unit that restores the intended grammatical relationship without introducing a new meaning.

If the error is subject–verb agreement, change the verb form rather than rewriting the clause. If the error is a preposition, change the preposition rather than replacing the whole phrase. If the error is a determiner, repair the determiner. Minimal edits are safer because every additional change creates another opportunity for error and another opportunity to distort meaning.

For example: Each of the experiments were repeated twice. The head controlling agreement is Each, singular. The minimal repair is was. A large rewrite such as The researchers repeated every experiment twice may be perfectly good English, but it no longer demonstrates the exact grammatical diagnosis and may not satisfy the task format.

Minimal repair is not rigidity. Sometimes the smallest successful unit is larger than one word. A broken comparative construction may require changing a phrase; a dangling modifier may require reattaching a clause; a reference problem may require repeating a noun. The principle is to avoid changes that are not justified by the identified failure.

False positives: the hidden cost of aggressive editing

Students often believe that good editing means making many corrections. That belief is dangerous. An editing mark scheme rewards correct decisions, not visible activity. A false positive occurs when a learner changes language that was already acceptable.

False positives have several causes. One is expectancy: if learners assume every line must contain an error, they force a change. Another is unfamiliarity: a less common but correct structure feels suspicious. A third is stylistic preference: the learner replaces a valid expression with a preferred alternative. A fourth is incomplete grammar knowledge: the learner applies a true rule outside the conditions where it is relevant.

Iona uses a simple burden-of-proof question: What exact grammatical relationship fails if I leave the sentence unchanged? If she cannot answer, she delays the edit. This does not mean never acting under uncertainty. It means uncertainty should be visible to the decision system.

Leonie adds a stop rule. If a sentence survives a meaning check, a structural check and the learner’s high-risk grammar checks, do not spend the entire paper trying to invent a fault. Mark it provisionally correct and return only if time remains.

Error family 1: subject–verb agreement

Subject–verb agreement is easy in short examples and much harder in authentic passages because the subject and verb are often separated by distracting nouns, relative clauses or prepositional phrases.

The most useful question is not “Which noun is nearest the verb?” but “What is the grammatical head of the subject?” In The quality of the photographs has improved, the head is quality, singular. In The photographs in the archive have faded, the head is photographs, plural.

High-risk patterns include each of, every one of, one of, collective nouns, quantities, titles, subjects joined by and, subjects joined by either…or or neither…nor, inverted structures beginning with there, and relative clauses whose subject is easy to misidentify.

Example: One of the reasons for the delays were a shortage of trained staff. The subject head is One; change were to was. Example: There appears to be several causes. In formal edited English, the notional subject after be is plural; appear is expected in many exam traditions. But because usage conventions vary, learners should follow the standard of the target assessment and avoid turning every variation into an absolute universal rule.

Error family 2: tense and aspect

Tense errors are rarely solved well by scanning for past or present forms in isolation. Build a timeline. Identify the reference time. Ask whether the event is completed, ongoing, repeated, prior to another event, or relevant to the present.

Example: She has submitted the form yesterday. The explicit finished past-time expression yesterday normally calls for simple past: submitted. Example: By next June, the team completes the survey. The future completion before a deadline is more naturally will have completed. Example: When the bell rang, we are discussing the final question. The ongoing past action should usually be were discussing.

Aspect matters because it changes how an event is viewed. The difference between worked, was working, has worked, has been working and had worked is not merely form. It encodes temporal structure. Editing therefore requires the learner to reconstruct that structure quickly.

Error family 3: auxiliary and verb-form chains

Many errors occur not in the tense choice itself but inside a multi-word verb phrase. After a modal, the lexical verb normally takes its base form: can go, not can goes. After have used as a perfect auxiliary, the past participle is required: has gone, not has went. In a progressive construction, be is followed by an -ing participle: is running. In a passive construction, be is followed by a past participle: was completed.

Editing improves when students read these as chains. Do not inspect the target word alone. Inspect the controlling auxiliary to its left and the grammatical job of the whole phrase.

Example: The results may indicates a change. May controls a base verb; indicates becomes indicate. Example: The bridge was build in 1988. Passive was requires the past participle built. Example: Researchers have began a second trial. Perfect have requires begun.

Error family 4: word form

Word-form errors test the architecture of the sentence. The learner may know the root word but fail to supply the form required by the slot: noun, verb, adjective or adverb.

In The rapid of the change surprised everyone, the determiner the and the following of-phrase signal a noun slot; rapid should become rapidity or, more naturally depending on context, speed. In The policy was widely success, the complement after was requires an adjective: successful. In She responded calm to the accusation, the verb responded is modified by an adverb: calmly.

Do not teach this as a suffix lottery. First identify the grammatical slot. Then select a form that fits both grammar and meaning.

Error family 5: articles and determiners

Articles are difficult because they encode definiteness, countability and discourse status. A learner who asks only whether a noun is singular will miss much of the system.

With singular count nouns, some determiner is normally required: a result, the result, this result, my result. Indefinite a/an typically introduces one nonspecific member of a class; the identifies a recoverable or unique referent. Zero article appears with many plural or uncountable nouns used generically: Books can change lives; Water expands when it freezes.

Editing must also respect fixed geographical, institutional and idiomatic patterns, which vary. Rather than memorizing hundreds of disconnected article rules, build the decision: countability → number → specific or generic reference → established expression.

Example: She gave me useful advice about the interview. Advice is uncountable in standard English, so no indefinite article is needed. Example: He bought laptop before the trip. Singular count noun laptop requires a determiner; if first mention and nonspecific, a laptop is likely.

Error family 6: prepositions

Preposition errors can look arbitrary because many are lexically selected: interested in, responsible for, depend on. Others express a semantic relationship: time, place, direction, instrument, cause, comparison, source or association.

The fastest route is to ask whether the preposition is controlled by a word or by the underlying relation. In They discussed about the proposal, the verb discuss takes a direct object in standard edited English; delete about. In She is capable to finish the task, the adjective capable selects of followed by a gerund phrase: capable of finishing. In The train arrived to the station, standard English normally uses arrived at for a point destination.

Preposition work benefits from phrase-level learning rather than single-word lists. Record the whole relationship in the error log.

Error family 7: pronouns and reference

Pronoun editing has three questions: agreement, grammatical case and reference clarity.

Agreement asks whether number and, where relevant, person are appropriate. Case asks whether the pronoun is acting as subject, object or possessive form. Reference asks whether the reader can identify the antecedent without guessing.

Example: Between you and I, the plan is risky. After the preposition between, standard formal English uses object case: me. Example: The company revised their policy may be acceptable in varieties that treat collective nouns as plural, but many assessments expect consistency with the chosen convention. Example: When Maya spoke to Elena, she was upset is structurally grammatical but referentially ambiguous if context does not resolve she. If an editing task targets grammar only, learners must be careful not to “correct” every stylistic ambiguity unless the task permits it.

Error family 8: modifiers

Modifiers must attach to the element they logically describe. Errors often arise when an introductory participial phrase lacks a plausible subject.

Example: Walking through the museum, the paintings impressed Leonie. Grammatically, the sentence makes the paintings the understood walker. The intended walker is Leonie. A suitable repair is Walking through the museum, Leonie was impressed by the paintings.

Single-word modifiers can also be misplaced. She almost answered every question correctly can mean she nearly answered them but perhaps did not. She answered almost every question correctly means she answered nearly all of them. Position changes meaning, so editing must identify what the modifier is intended to limit.

Error family 9: parallel structure

When a sentence presents elements as equivalent, their grammatical forms should usually align. Parallelism errors often hide inside lists, comparisons and paired conjunctions.

Example: The course teaches students to analyse evidence, writing clearly and how to evaluate sources. The three complements are structurally mismatched. A parallel repair could be to analyse evidence, to write clearly and to evaluate sources, or analysing evidence, writing clearly and evaluating sources, depending on the larger sentence.

With not only…but also, either…or, neither…nor and comparisons using than or as, mark the paired elements and compare their grammatical shapes.

Error family 10: conjunctions and logical relationships

Conjunctions encode logic. A sentence can be grammatically well formed at the clause level yet wrong because the connector misrepresents the relationship.

Example: Although the roads were flooded, therefore the match was postponed. The sentence double-marks the concessive/cause relationship in a way many formal exam systems reject. One repair is Because the roads were flooded, the match was postponed; another is The roads were flooded; therefore, the match was postponed. Which is minimal depends on the identified target.

Do not swap connectors because one sounds more formal. Name the relation: cause, contrast, concession, condition, addition, example, sequence, result or alternative. Then select the form that encodes it.

Error family 11: countability and number

Noun errors often involve a hidden question: is the noun countable in the intended meaning?

Standard English treats nouns such as information, advice, equipment and furniture as uncountable in their common senses. That affects articles, plural forms and quantifiers. Many informations is therefore wrong; much information or many pieces of information may work. By contrast, some nouns switch between count and noncount meanings: paper as material and a paper as an article or examination.

Number also interacts with agreement and reference. Do not repair a plural noun in isolation without checking what it controls elsewhere in the sentence.

Error family 12: comparison

Comparisons fail when the forms compared are grammatically or logically mismatched.

Example: The rainfall in City A is higher than City B. Strictly, rainfall is being compared with a city. A clearer structure is The rainfall in City A is higher than that in City B. Example: This method is more easier double-marks the comparative; use easier. Example: No other runner was as faster as Maya mixes two comparison patterns; use as fast as.

Editing comparisons requires both grammar and logic. Ask: what two things are being compared, and are they expressed in parallel forms?

Error family 13: finite and non-finite verb patterns

Many lexical items select an infinitive, a gerund, or another clause pattern. Learners often know the words but not the constructions.

Decide to go, avoid going, suggest going, make someone go, allow someone to go, look forward to going. The final example is especially useful because to is a preposition rather than an infinitive marker, so it is followed by a noun-like -ing form.

Rather than memorizing as isolated “gerund versus infinitive” rules, store the verb or adjective with its complement frame. In an editing passage, identify the controlling word, then test whether the following form matches its frame.

Error family 14: relative clauses

Relative clauses introduce errors in pronoun choice, agreement, unnecessary repetition and punctuation.

Example: The student who she won the prize thanked her teacher. Who already functions as the subject of won; delete she. Example: The books that is on the shelf belong to Iona. The relative pronoun takes its number from antecedent books; change is to are. Example: My brother, who lives abroad is visiting needs the closing comma if the clause is non-restrictive in the target punctuation convention.

When editing, mark the antecedent, the grammatical role of the relative word inside the clause and whether the clause identifies or merely adds information.

Error family 15: sentence boundaries

Some editing tasks include fragments, run-ons or comma splices; others restrict errors to single-word grammar. Know the task. Where sentence boundaries are testable, identify independent clauses before deciding on punctuation or conjunction.

The experiment failed, the team repeated it. Two independent clauses cannot normally be joined by a comma alone in formal edited English. Repair with a full stop, semicolon, coordinating conjunction or suitable subordination depending on meaning. Because the experiment failed. is a dependent fragment if presented as a complete sentence without a main clause.

Boundary editing is high value because a wrong repair can damage the entire clause structure. Do not treat punctuation as decoration.

Error family 16: lexical choice versus grammar

Not every strange word is a grammar error. Some tasks test vocabulary or collocation; others explicitly do not. Learners need a boundary.

Heavy rain is conventional; strong rain may be understandable but collocationally less standard. Make a decision and do homework are conventional pairings. Yet a grammar-only task might not ask you to replace every unusual collocation. Read the instructions and examples from the target assessment.

When the scope is uncertain, prioritize relationships that are demonstrably grammatical rather than stylistic. This protects against over-editing.

Correct lines are active decisions

A correct line is not a line you “could not find an error in.” It is a line that survived a reasonable set of tests.

Leonie uses a three-pass acceptance rule when the format includes correct items. First, read for complete meaning. Second, check the highest-risk structural relationships: subject–verb, verb chain, tense timeline, determiner–noun, pronoun reference and selected complement patterns. Third, ask whether any proposed change is supported by a specific rule rather than preference. If the line survives, accept it and move.

This matters because learners who cannot confidently mark a line correct waste time and create errors. The skill being examined is discrimination, not suspicion.

Passage context changes the answer

Sentence correction exercises are useful early, but passage editing is the real transfer test. Context affects tense, reference, definiteness, lexical meaning and cohesion.

Suppose a passage begins: Last year, a research team began a long-term study of urban birds. Later, a sentence reads: The scientists collect data every morning for six months. If the six-month period belongs to last year, collected is likely. But if the sentence shifts into a general description of the ongoing protocol, present tense might be appropriate. The editor must track discourse time.

Similarly, a device may be correct at first mention, while the device is expected once the referent has been established. Articles are passage-level signals, not just noun-level decorations.

A practical scanning order

There is no universally best scan, but a useful default for many secondary learners is:

1. Read the whole sentence for meaning.
2. Find finite verbs and their subjects.
3. Check auxiliary–verb chains and the time line.
4. Check nouns for number, countability and determiner support.
5. Check pronouns and references.
6. Check adjective/adverb and other word-form slots.
7. Check selected prepositions and complement patterns.
8. Check modifiers, comparisons and parallel structures.
9. Only then inspect lower-priority personal risks.

The order should be personalized. If a learner almost never misses articles but frequently misses non-finite verb patterns, move the latter earlier. A mature editor owns a risk-weighted scan, not a generic checklist.

Why random rereading underperforms

Random rereading feels diligent because the eyes move across every word. It is weak because attention has no target. Working memory becomes saturated by meaning, grammar, spelling, punctuation and uncertainty at the same time.

Structured passes reduce that load. Maren may first mark clause boundaries and heads. Iona may then test the suspect relationship. Leonie limits the time spent on low-confidence items. Over training, these passes become faster and partially automatic. The goal is not to perform a visible nine-step ritual forever; it is to develop reliable attention habits that compress under examination pressure.

The confidence ladder

Every proposed edit can be assigned one of three confidence levels.

• High confidence: the structure clearly violates a known relationship and the correction is unambiguous.
• Medium confidence: there is a likely error, but more than one analysis or repair is plausible.
• Low confidence: the sentence feels odd, but the learner cannot state a grammatical reason.

Handle high-confidence items immediately. Mark medium-confidence items for a quick return if the format allows. Avoid spending early time on low-confidence stylistic hunches. This is not merely time management; it is error control.

Verification after correction

A correct diagnosis can still produce a wrong final answer if the replacement is malformed. Always reread the repaired sentence.

Verification asks five questions: Does the new form fit the grammatical slot? Does it agree with controlling words? Does it preserve the intended time? Does it preserve the intended meaning? Did the change create a new error elsewhere?

Example: Neither of the explanations are convincing. A learner correctly diagnoses agreement and changes are to is. Rereading confirms Neither…is. Example: She suggested to take the earlier train. A learner recognizes a complement error and changes to take to taking. Rereading confirms the whole frame suggested taking. The extra two seconds are cheaper than losing a mark to an incomplete repair.

Error logs should record causes, not just corrections

An error log that records only “were → was” is weak. It captures the answer but not the transferable structure.

A useful entry contains: original fragment; correction; error family; controlling relationship; reason the error was missed; confidence before checking; and a new example made by the learner.

For The group of volunteers were ready, the log might say: Family: subject–verb agreement. Controller: head noun group. Why missed: attracted by nearby plural volunteers. Transfer cue: when a verb follows an of-phrase, check the noun before of first.

That cue can travel into unseen passages. The correction alone cannot.

Detection rate and correction rate are different metrics

Suppose a learner is given twenty seeded errors. She independently notices twelve and correctly repairs eleven of those. Her detection rate is 60 percent; her correction rate conditional on detection is about 92 percent. The main bottleneck is not grammar repair. It is search.

Another learner notices eighteen but correctly repairs only ten. That learner has strong detection but weak diagnosis or correction knowledge.

These profiles should not receive the same homework. The first needs mixed detection drills with unlabelled targets. The second needs narrow grammar reconstruction and explanation. Measurement prevents generic practice.

False-positive rate belongs beside accuracy

Imagine two learners both correct eight genuine errors. Learner A changes no correct lines. Learner B changes four correct lines as well. Their raw “errors found” count hides a large difference in judgment.

Track false positives explicitly: correct lines changed ÷ correct lines presented. As proficiency rises, the learner should not only find more genuine errors but also disturb fewer correct structures.

This is the editing equivalent of precision in information retrieval. High recall without precision means the learner finds many real errors but also flags too much valid language. World-class examination performance requires both.

Time pressure changes cognition

Under time, learners rely more heavily on habit and pattern recognition. Long conscious reasoning becomes expensive. That is why the training sequence must begin slowly enough to build accurate models, then compress.

Leonie separates learning speed from performance speed. During learning, explain why. Draw the clause structure. Mark the subject head. Build the timeline. During performance, compress that reasoning into a cue: head noun? time anchor? modal + base? determiner + singular count noun? pronoun antecedent?

If speed is imposed before the reasoning exists, the learner simply becomes faster at guessing.

Timed scanning without panic

A useful timed routine for a short editing passage is:

1. Read once for overall meaning and time frame.
2. Work line by line, but keep sentence boundaries visible; a sentence may cross lines.
3. Take high-confidence corrections immediately.
4. Place a light mark beside uncertain lines rather than freezing.
5. Complete the entire first pass.
6. Return to uncertain lines with remaining time.
7. Use the final seconds to verify answer transfer, spelling of corrections and whether any line was skipped.

The exact minutes depend on the paper. The principle does not: protect completion, reserve a small review window, and do not let one ambiguous item consume the whole task.

Line breaks are not grammar boundaries

Many editing formats print one line at a time. Learners then start treating the line break as if it were a sentence break. This causes errors.

A subject may occur on one line and its verb on the next. A relative clause may open in one line and close later. A pronoun may refer to a noun several lines above. The unit of grammar is the clause or phrase, not the typographic line.

Maren’s rule is to keep a mental “syntax bridge” across line endings. Before changing a word, read far enough left and right to identify what controls it.

One-error-per-line formats create a special trap

If an examination guarantees at most one grammatical error per tested line, that is useful information but should not become a mechanical shortcut. Finding one genuine error increases confidence that the rest of the line can be left, yet the diagnosis still has to be correct. Conversely, if the format includes some error-free lines, the learner must preserve the possibility of zero.

Training should therefore vary the number and position of errors. If every practice line always contains exactly one obvious mistake, learners can succeed by expectancy rather than grammar. Authentic skill appears when the distribution is less predictable.

Why answer keys need explanations

An answer key that says only were → was is sufficient for marking but weak for learning. A useful answer explains the relationship: Each is the singular head of the subject, so the finite verb must be singular.

Explanations also expose ambiguity. Sometimes more than one grammatical rewrite is possible. A strong key identifies the minimal intended repair and, where appropriate, explains why another tempting answer changes meaning or exceeds the task.

The first spoke of this lane therefore treats answers as micro-lessons, not just solutions.

Practice should move from isolated to integrated

A reliable progression has five stages.

1. Single-family recognition: only subject–verb agreement, only tense, only word form.
2. Single-family detection: mix correct and incorrect items inside the same family.
3. Mixed sentence editing: several families compete.
4. Passage editing: context controls tense, articles and reference.
5. Timed unseen simulation: unfamiliar topic, authentic density, no family labels.

Skipping from rule explanation straight to full passages overwhelms some learners. Remaining forever in isolated drills produces brittle skill. The training path must bridge them.

Interleaving beats blocked comfort

Blocked practice feels productive because twenty subject–verb items make the target obvious. Interleaving forces the learner to identify the target family before applying the rule.

A useful week may begin with blocked repair, then move quickly to mixed sets. The editing exam does not announce, “This line contains a preposition error.” It presents language. The learner must classify first.

Interleaving therefore trains the hidden first question: What kind of problem, if any, am I looking at?

Unseen topics are part of the skill

An editor who performs only on familiar school topics has not yet demonstrated transfer. Passages should range across science, history, environment, technology, sport, culture, everyday life and unfamiliar domains.

The content does not need specialist knowledge to solve the grammar, but unfamiliar vocabulary increases cognitive load. Training under varied topics teaches the learner to separate what must be understood for grammar from what can remain background.

Iona asks: what minimum meaning is needed to resolve the grammatical relationship? That keeps unfamiliar content from hijacking the entire task.

Reading difficulty and editing difficulty can interact

If a learner cannot parse the sentence, grammar editing becomes harder. That does not mean the learner needs to study every word. It means sentence comprehension and editing are partly coupled.

Long noun phrases, embedded clauses and passive constructions raise parsing demands. Advanced editing practice should therefore include syntactically complex but semantically accessible passages. The goal is to build structural reading, not obscure vocabulary trivia.

Personal risk maps

A personal risk map ranks error families by four variables: frequency, severity, detectability and repairability.

An error that occurs often, costs marks and can be fixed reliably once noticed should be high priority. An exotic construction that appeared once may not deserve the same training load. The map should be updated from real work, not from a generic list of “common grammar mistakes.”

Maren may need to watch modifiers and clause boundaries because she builds complex sentences. Iona may need to watch over-analysis that turns acceptable alternatives into false positives. Leonie may need to watch tense endings and number agreement late in the task when speed rises. The same syllabus can produce different risk profiles.

Editing under fatigue

Accuracy often declines near the end of a paper. Small inflectional errors become harder to see. Attention narrows. Learners either rush or reread without focus.

Train some editing sets after another demanding English task rather than always at the beginning of a fresh session. This tests whether the routine survives fatigue. The goal is not to exhaust students for its own sake. It is to verify that the final scan is robust when cognitive resources are no longer perfect.

Recovery after an uncertain item

One hard line can contaminate the next five if the learner carries frustration forward. Recovery is therefore part of editing performance.

Leonie’s reset is brief: mark the uncertainty, release it, take one complete breath if needed, then read the next sentence from meaning rather than from panic. The learner can return later. This is a task-control technique, not a relaxation ritual. It protects working memory from unresolved rumination.

When two corrections seem possible

First ask whether both are genuinely grammatical. Second ask which changes less. Third ask which better preserves the passage meaning. Fourth ask whether the task expects a single-word correction or permits restructuring.

Suppose the sentence is The teacher recommended the students to revise the chapter. In some varieties and registers, this construction may be judged differently. A safer formal rewrite in many assessment contexts is recommended that the students revise or recommended revising, but a one-word correction task may signal that a different target was intended. This is why learners must work within the assessment’s conventions and not universalize every answer beyond context.

Standard English and variation

English is global. Grammatical conventions vary across regions and registers. Collective-noun agreement, certain prepositions, punctuation choices, pronoun usage and lexical patterns can differ while remaining legitimate in recognized varieties.

An examination still needs a marking standard. The sensible approach is to teach the target standard explicitly while acknowledging that “different from the mark scheme” is not always the same as “ungrammatical everywhere.” This distinction improves linguistic accuracy and prevents students from learning brittle myths.

Editing and vocabulary knowledge

Grammar editing sometimes depends on knowing how a word behaves. You need to know that responsible commonly takes for, that suggest does not normally take a person + infinitive pattern in standard formal English, or that advice is uncountable in its common sense.

Vocabulary study for editing should therefore include grammar around the word: part of speech, countability, complement patterns, common collocations and derived forms. This is more useful than memorizing isolated definitions.

Editing and sentence parsing

Advanced performance depends on seeing the skeleton of a sentence quickly. You do not need to label every constituent with university-level terminology. You do need to answer practical structural questions:

• Where is the main clause?
• Which noun phrase is the subject?
• What is its head?
• Which finite verb belongs to it?
• Which clauses are embedded?
• What does this modifier attach to?
• What element does this pronoun replace?

Maren treats these questions as X-rays. Once the skeleton is visible, many errors stop being mysterious.

Editing as constrained search

A useful technical model treats editing as a search process over a large language space. At every word, countless alternatives are possible, but only a few are relevant. Grammar knowledge reduces the search space.

If you identify a modal, you know the following verb is likely to be base form. If you identify a singular count noun with no determiner in a position where one is required, the candidate repairs narrow. If you identify a comparative than-phrase, you test parallelism. Expertise is partly the ability to generate fewer, better hypotheses.

This is why random synonym replacement is not editing. It expands the search space instead of constraining it.

A decision matrix for a suspect word

Question	If yes	If no
Does the sentence meaning make sense?	Continue to structural testing.	Read wider context before editing.
Can I name the controlling relationship?	Test the rule.	Do not change on instinct alone.
Is the original incompatible with that relationship?	Generate a repair.	Preserve the original.
Does the repair preserve meaning?	Verify locally.	Reject and reconsider.
Does the repaired sentence satisfy the task format?	Record it.	Choose a format-compliant repair.

Diagnostic set A: agreement and attraction

Try these before reading the explanations.

1. The collection of old maps were donated to the library.
2. Each of the finalists have prepared a short speech.
3. The effects of the new policy is still being studied.
4. Neither the coach nor the players was satisfied with the result.
5. The series of workshops begins next month.

Answers. 1: were → was; head noun collection. 2: have → has; Each is singular. 3: is → are; head noun effects is plural. 4: in conventions using proximity agreement with neither…nor, was → were because the nearer subject is players; learners should follow the target assessment’s rule. 5: no change; series is singular here and begins agrees.

The fifth item is essential. If your practice never contains correct examples, you cannot measure discrimination.

Diagnostic set B: verb chains and time

1. The committee has chose a new chairperson.
2. By the time the doors opened, hundreds of people waited outside.
3. Leonie can completes the first scan in four minutes.
4. The samples were store at a constant temperature.
5. Since January, the team has collected data every week.

Answers. 1: chose → chosen; perfect auxiliary has requires a past participle. 2: if the waiting began before the doors opened and continued up to that point, had been waiting is the clearest intended relationship; a simpler were waiting may fit a different nuance, so context matters. 3: completes → complete; modal + base form. 4: store → stored; passive construction. 5: no change; present perfect with since January is well motivated if the period reaches the present.

Diagnostic set C: word form and complement

1. The new method is remarkable efficient.
2. The announcement caused widespread confuse.
3. Iona insisted to checking the figures again.
4. The students avoided to make the same mistake.
5. The explanation was surprisingly clear.

Answers. 1: remarkable → remarkably; an adverb modifies adjective efficient. 2: confuse → confusion; noun slot after adjective widespread. 3: a standard repair is insisted on checking; the controlling pattern is insist on + -ing. 4: to make → making; avoid takes a gerund complement. 5: no change.

Diagnostic set D: determiners, countability and prepositions

1. She gave us an useful information before the test.
2. The report provides several evidence for the claim.
3. Maren is interested on how sentences are built.
4. The team discussed about the final schedule.
5. The equipment was delivered on time.

Answers. 1: remove an; information is normally uncountable. 2: several evidence is incompatible with the uncountable noun in standard usage; use several pieces of evidence if rewriting is allowed. 3: on → in; interested in. 4: remove about; discuss takes a direct object in standard formal English. 5: no change.

Diagnostic set E: reference, modifier and parallelism

1. After reading the passage, the answer became obvious to Iona.
2. The programme aims to improve accuracy, increasing speed and to build confidence.
3. When Maren met Iona after the paper, she looked relieved.
4. The new scanner is faster than the old model.
5. Walking across the hall, Leonie noticed the clock.

Answers. 1: dangling modifier; a repair is After reading the passage, Iona found the answer obvious. 2: restore parallel forms, for example to improve accuracy, increase speed and build confidence. 3: grammatical but potentially ambiguous; do not automatically alter it in a grammar-only single-error task unless reference clarity is within scope. 4: no change. 5: no change; Leonie is the understood walker.

Why some “obvious” corrections are wrong

Students often overapply familiar classroom rules. “Never use passive voice” is not a grammar rule. “Never start with And” is not a universal grammar rule. “Never split an infinitive” is not a reliable modern prohibition. “A sentence cannot end with a preposition” is not a useful universal rule. Examination editing should be grounded in current standard grammar and the specific marking conventions of the assessment, not inherited folklore.

When a correction makes the sentence stiffer without solving a structural problem, question whether it is needed.

Build explanations in one sentence

For training, require each correction to be justified in one compact sentence. Examples:

• was → were because the head of the subject noun phrase is plural.
• went → gone because has requires a past participle.
• quick → quickly because the word modifies the verb responded.
• to → for because the adjective responsible selects for.
• a → the because the noun refers back to a specific item introduced earlier.

If the explanation cannot be stated, the learner may have guessed. Over time the spoken explanation can disappear; the mental relationship remains.

The role of retrieval

Editing is not helped much by rereading grammar notes passively. The learner needs to retrieve the relationship when the passage does not announce it.

Short retrieval prompts work well: What follows a modal? What form follows perfect have? How do you find the subject head? Which common nouns are uncountable? What complement follows avoid? What does a dangling modifier attach to?

Then put those relationships back into mixed passages. Retrieval without application becomes quiz knowledge; application without retrieval becomes guessing.

Spacing and forgetting

An error family that is mastered on Monday can disappear by Friday. Revisit important patterns across weeks rather than exhausting them in one session. Spaced practice reveals whether the relationship has entered durable memory.

A useful schedule for a recurring error might be: same day explanation; two days later mixed sentence; one week later passage; two weeks later unseen timed passage; one month later maintenance check. The exact intervals can vary. The principle is to make forgetting visible before the examination does.

Practice density

If every line in a worksheet contains an error, the learner is not practising the decision “correct or leave.” If errors are too rare, the learner may not get enough repetitions. Vary density.

Early drills can be dense to build the target pattern. Mixed sets should include correct items. Advanced passages should resemble the uncertainty of the actual task. This is one way to prevent overfitting.

Seeded-error design

Teachers can construct strong exercises by starting with a clean passage and seeding errors deliberately. Keep a ledger of the inserted family, the intended correction and the reason. Avoid accidental second errors in the same sentence unless the task is designed for them.

Seed errors that arise from plausible learner confusions rather than absurd mistakes. The results shows is realistic. A sentence that is nonsensical merely to force a correction teaches less.

After the learner completes the passage, compare not just answers but search behaviour. Which families were invisible? Which correct lines were disturbed? Which wrong forms were noticed immediately?

Using authentic learner writing carefully

Real learner writing contains naturally occurring errors, but it differs from a supplied-text editing task because there may be multiple errors, stylistic weaknesses and ambiguous intent. It is still useful for diagnosis, provided the teacher labels the purpose.

One exercise can ask: find only subject–verb agreement errors. Another can ask: identify every grammar issue you can justify. A third can ask: make the smallest corrections while preserving the writer’s meaning. These tasks build transfer without pretending that authentic writing follows an exam’s controlled distribution.

Mock editing papers

A mock should test more than grammar. It should reproduce the cognitive conditions of the target: passage length, line layout if relevant, proportion of correct items, topic unfamiliarity, answer method and time constraint.

After the mock, separate score into components: genuine errors detected, genuine errors repaired correctly, correct lines preserved, time used, items left blank, and late-stage accuracy. That produces a performance profile rather than one total mark.

The lane’s dedicated mock-paper spoke will deepen this system.

Score stability

One high score is not yet stable skill. Editing performance is stable when accuracy remains within a narrow range across different topics, error mixes and days.

Track at least several sets. If marks swing widely, identify what changes with the low scores: unfamiliar syntax, certain error families, time, fatigue, or false positives. Training should reduce variance, not only increase the best-case score.

The independence test

A learner is not ready simply because she can explain an answer after a teacher points to the error. Independence means she can detect, diagnose, repair and verify without cues.

Use four levels:

1. Prompted repair: the error is identified for the learner.
2. Family-cued detection: the learner is told the error family.
3. Mixed detection: no family cue, sentence-level items.
4. Unseen passage performance: mixed, contextual, timed and independently verified.

Progress through the levels. Do not mistake success at Level 1 for exam readiness.

A six-week repair programme

Week 1: map the baseline

Use two mixed untimed sets and one short timed passage. Record detection rate, correction rate, false positives and time. Build the first personal risk map. Do not rush to teach everything. Find the dominant bottleneck.

Week 2: rebuild the top two error families

Teach structure explicitly. Use blocked examples, then insert correct controls. Require one-sentence explanations. End with a mixed set containing the target families among distractors.

Week 3: expand to passage context

Use short passages where tense, reference and articles depend on previous sentences. Force the learner to read beyond the line. Begin using confidence ratings.

Week 4: compress the scan

Set moderate time limits. Keep accuracy as the primary target. Teach skip-and-return behaviour for medium-confidence items. Measure whether speed increases without false positives rising.

Week 5: unseen interleaving

Use unfamiliar topics and mixed syntax. Remove family labels. Include correct lines. Test after another English task once to expose fatigue effects.

Week 6: simulate and retest

Run authentic-condition editing tasks. Compare with Week 1 on all metrics. Keep only the highest-value personal scan cues for the final routine.

A twelve-week mastery programme

For learners with more time, extend the system: Weeks 1–2 diagnose and rebuild agreement/verb chains; Weeks 3–4 add determiners, countability, prepositions and word form; Weeks 5–6 add modifiers, reference, comparisons and parallelism; Weeks 7–8 shift heavily into passage context and correct-line discrimination; Weeks 9–10 compress under time and vary topics; Week 11 runs multiple mocks; Week 12 focuses on stability, recovery and maintenance.

The schedule is not a fixed syllabus. If the learner’s data shows a different bottleneck, change the load. Training should follow evidence.

Teacher feedback that transfers

Weak feedback says, “Wrong. Use was.” Better feedback asks, “What is the head of the subject?” The learner then reconstructs the relationship.

For a missed error, identify whether the learner did not see the structure or did not know the rule. For a false positive, ask what rule the learner thought was being applied. For a wrong correction after correct detection, compare alternative repairs. The goal is to make the learner’s decision process visible.

This links directly to English Exam Feedback Transfer: a correction should change future behaviour, not merely fix yesterday’s line.

Parent or tutor diagnostic questions

• Can the learner explain why an answer is wrong without seeing the key?
• Does the learner miss errors or misrepair errors that were detected?
• Does accuracy fall mainly in passages rather than isolated sentences?
• Does the learner change correct lines?
• Which three error families account for most losses?
• Does performance collapse under a clock?
• Does the learner reread randomly or use a structured scan?
• Can the learner recover after one difficult item?
• Is the error log organized by transferable relationships?
• Are recent scores stable across unseen topics?

These questions produce a more useful plan than “grammar is weak.”

Advanced passage: structural load

Read the passage and identify the grammatical problems. Some sentences are correct.

Over the past decade, a network of community laboratories have expanded across the region. Each of the laboratories provide local schools with access to equipment that would otherwise be too expensive to purchase. The programme began when a group of engineers were looking for a way to share unused instruments. Since then, hundreds of students have took part in weekend workshops. The organizers say that the most valuable outcome are not the experiments themselves but the confidence students gain from designing them. To keep the programme sustainable, every laboratory is ask to train new volunteers. The network, which now includes thirty sites, continues to grow.

Analysis. network…have → has; the subject head is network. Each…provide → provides. have took → have taken. outcome are → outcome is because the subject is singular; the complement after but does not control the verb. is ask → is asked for the passive. a group of engineers were looking may be acceptable under plural notional agreement in some varieties and singular in others; the target standard matters. The final sentence is correct.

Notice the method: head nouns, auxiliary chains and passive form solve most of the passage. It is not necessary to “check every grammar rule.”

Advanced passage: time and reference

When Iona entered the archive last Tuesday, the technician already finishes scanning the first box of documents. He explained that the files were being digitized because several originals has begun to deteriorate. The archive introduced the project two years earlier, and since then its staff had processed more than ten thousand pages. By the end of this year, they expect to completed the remaining collection. Iona asked whether the digital copies would be available to the public, and the technician said that they would.

Analysis. already finishes → had already finished because the completion precedes another past event. originals has → originals had or have depending on the intended reporting-time relationship; in this narrated past context, had begun is strongly motivated. since then…had processed can be correct if the reference point remains in the narrated past; if the passage shifts to the present, have processed would be needed. expect to completed → expect to complete. The final pronoun they refers to the digital copies and is plausible.

This passage shows why tense cannot be edited one verb at a time. The timeline is the system.

Advanced passage: forms, modifiers and complements

Maren joined a student research group because she was curious about how public spaces influence behaviour. The team planned to observe how people moved through a busy square and to compare their findings with data from a quieter park. Before beginning, the supervisor reminded them to record observations systematic. He also advised that they should avoid to make assumptions about why a person chose a particular route. Walking through the square, the sound of traffic made concentration difficult. Despite the challenge, the students worked careful and produced a surprisingly detail set of notes.

Analysis. systematic → systematically; adverb modifies record. avoid to make → avoid making. Walking through the square, the sound of traffic… is a dangling modifier; make the students the subject, e.g. Walking through the square, the students found that the sound of traffic made concentration difficult. worked careful → worked carefully. detail set → detailed set. The parallel to observe…and to compare is correct.

A 20-item mixed calibration

1. The number of applications have increased this year.
2. Every student in the two classes were given a copy.
3. The results suggest that the new method is effective.
4. She has wrote three reports since January.
5. By noon, the technicians repaired the fault and left.
6. He is capable to solve the problem independently.
7. The researchers conducted the experiment careful.
8. There are less errors in the second draft.
9. Neither of the explanations seem convincing.
10. The information were stored securely.
11. The committee discussed the proposal for an hour.
12. The students were asked to submit their work by Friday.
13. After finishing the test, the bell rang.
14. The new policy aims to reduce waste, improving efficiency and to lower costs.
15. She is responsible of checking the final figures.
16. The equipment arrived yesterday and was installed immediately.
17. Each of the machines need regular maintenance.
18. The team has been working on the project for six months.
19. Maren explained the rule to Iona, which was confusing.
20. The evidence supports the conclusion.

Calibration key. 1 have → has; head number. 2 were → was; Every student singular. 3 correct. 4 wrote → written. 5 can be correct if both actions were completed before noon; do not force a perfect merely because by appears. 6 to solve → of solving. 7 careful → carefully. 8 in formal count-noun usage, less → fewer. 9 seem → seems under singular Neither convention. 10 were → was; information uncountable singular agreement. 11 correct. 12 correct. 13 dangling modifier; the bell did not finish the test. 14 parallelism; e.g. to reduce waste, improve efficiency and lower costs. 15 of → for. 16 correct. 17 need → needs. 18 correct. 19 faulty or ambiguous relative reference; restructure if reference clarity is within scope. 20 correct.

Score more than the total. Count genuine errors missed, false positives and the families involved.

The five-second proof

As learners become fluent, many edits can be validated with a short internal proof:

• Agreement: “Head is singular, so verb singular.”
• Perfect: “Have + past participle.”
• Modal: “Modal + base.”
• Word form: “This slot modifies a verb, so adverb.”
• Determiner: “Singular count noun needs a determiner.”
• Complement: “Avoid + -ing.”
• Preposition: “Responsible for.”
• Modifier: “Who is doing the -ing action?”

This is compressed reasoning, not memorized magic. Each proof is backed by a structure learned earlier.

When to use a grammar reference

During training, verify uncertain constructions in a reputable contemporary grammar reference or dictionary. Do not build an error log from guesses. Dictionaries are especially useful for countability, complement patterns and collocations because they show how a word behaves in real syntax.

During an examination, that resource may be unavailable. The purpose of reference checking beforehand is to improve the learner’s internal model.

What not to do the week before an examination

Do not suddenly introduce fifty rare grammar rules. Do not replace the learner’s scan routine with a completely new checklist. Do not spend all revision time on untimed worksheets. Do not infer ability from one excellent set. Do not chase every stylistic preference.

Instead, review the personal risk map, perform several mixed passage sets, inspect false positives, rehearse the timing routine, and keep explanations tied to relationships the learner already understands.

The day-before routine

One short mixed set is enough for many learners. Review the final cues: read meaning, find the subject head, inspect the verb chain, check time anchors, look at determiner–noun relationships, verify pronoun reference, test known complement patterns, preserve correct lines, move on after the evidence is sufficient.

The goal is readiness, not exhaustion.

The examination-room routine

Start with the passage, not with panic. Establish topic and time frame. Work at sentence level even if the page is divided into lines. Take clear corrections. Park uncertainty. Protect correct lines. Keep handwriting or answer transfer unambiguous. Return to marked items. Reread every repair in its sentence. Stop when further changes are based on discomfort rather than evidence.

What improvement looks like

Early in training, the learner may say, “It sounds wrong.” Later, she says, “The head noun is singular but the finite verb is plural.” Later still, she sees the relation almost instantly and corrects it without needing to verbalize the full explanation.

At the same time, false positives fall. Passage accuracy approaches sentence accuracy. Time becomes predictable. Scores vary less across topics. The learner recovers after uncertainty instead of spiralling. These are better indicators of mastery than the number of grammar rules memorized.

The SETC Secondary Editing route

This hub will grow as a tightly separated series. Each spoke has one job so that the library gains depth without turning into duplicate pages.

• SETC-SEC-EDIT-010: English Error Correction Exercises With Answers | Train Detection Before Repair.
• SETC-SEC-EDIT-020: Grammar Editing Exercises | Subject–Verb Agreement, Tense, Word Form and Prepositions.
• SETC-SEC-EDIT-030: Passage Editing Practice | Find Grammar Errors in Context Without Changing Meaning.
• SETC-SEC-EDIT-040: English Editing Under Time Pressure | Scan, Verify and Stop Without Guessing.
• SETC-SEC-EDIT-050: Sentence Correction Exercises | Diagnose the Grammar Relationship Before Changing a Word.
• SETC-SEC-EDIT-060: Editing Error Types | Build a Personal Grammar Error Map.
• SETC-SEC-EDIT-070: Editing Accuracy | Avoid False Positives and Unnecessary Changes.
• SETC-SEC-EDIT-080: Proofreading vs Editing for English Exams | Know Which Errors the Task Actually Tests.
• SETC-SEC-EDIT-090: Grammar Error Log | Turn Corrections into Transfer.
• SETC-SEC-EDIT-100: Mock Editing Papers | Simulate, Diagnose, Repair and Retest.
• SETC-SEC-EDIT-110: Advanced Grammar Editing | Embedded Clauses, Modifiers, Parallelism and Reference.
• SETC-SEC-EDIT-120: Editing Score Stability | Make Accuracy Repeatable Across Unseen Passages.

Connections to the wider SETC system

Editing is a narrow task, but it sits inside a larger performance ecology. English Language Control supplies the wider accuracy foundation. English Exam Error Analysis helps identify the first repairable cause. English Exam Practice Design governs the movement from drills to authentic transfer. English Mock Exam Performance tests the skill under realistic load. English Exam Feedback Transfer converts mistakes into future decisions.

For the wider Secondary route, return to Secondary English | Reading, Writing and Examination Performance. For editing your own completed composition rather than a supplied passage, use English Editing Performance | Detect, Prioritise and Repair Errors Under Test Conditions.

Final operating model

UNDERSTAND THE SENTENCE → FIND THE CONTROLLING RELATIONSHIP → TEST THE FORM → CHANGE ONLY WHAT THE EVIDENCE REQUIRES → REREAD THE REPAIR → PRESERVE CORRECT LANGUAGE → MOVE.

Maren’s contribution is structure: see what controls what. Iona’s contribution is evidence: prove the fault before changing the text. Leonie’s contribution is execution: make the decision fast enough to survive the whole paper. When those three lenses work together, editing stops being a guessing game. It becomes a disciplined form of grammatical reasoning.

Extended field manual: 40 high-value editing situations

1. Agreement after long prepositional phrases

When the subject is followed by an of-phrase, mentally bracket the modifier. In The condition of the engines was checked, engines is plural but sits inside the modifier. The head condition controls was. A useful exercise is to remove the modifier temporarily: The condition was checked. This stripping technique makes attraction errors visible. It is especially valuable when several nouns intervene: The condition of the engines in the older vehicles was… The number of intervening plurals does not change the head.

2. Agreement with quantifying expressions

Expressions such as a number of and the number of behave differently in standard usage. A number of students are absent treats the phrase as meaning several students; The number of students is rising has singular head number. Learners often memorize these as exceptions. A stronger approach is to inspect meaning and head structure. The first construction functions semantically like a plural quantifier; the second refers to one quantity.

3. Collective nouns

Words such as team, committee and family vary between singular and plural agreement across English varieties and intended meanings. A team can be conceptualized as one unit or as its members. Examinations generally standardize an expected convention. Learn that convention, but do not teach the alternative as universally wrong. Within a passage, consistency matters.

4. Indefinite pronouns

Forms such as each, everyone, someone, neither and either often take singular agreement in formal edited English. The trap occurs when a plural noun appears nearby: Each of the players has… Strip the of-phrase if needed. Contemporary pronoun reference can use singular they with many indefinite antecedents; learners should distinguish agreement in the finite verb from later pronoun choice.

5. Correlative subjects

With either…or and neither…nor, many formal styles use agreement with the nearer subject when the subjects differ in number. Because real usage can be more complex, assessments often prefer avoiding awkward mixed-number constructions in free writing. In editing, however, follow the target convention and inspect which noun is closest to the verb.

6. Existential there

In There are several reasons, there is not the notional plural subject. The postverbal noun phrase carries number. Informal speech often uses there’s before plurals, but many formal editing tasks expect there are. This is an example where spoken frequency and examination standard may differ.

7. Narrative tense chains

Do not force every verb in a past narrative into simple past. Background actions may be progressive, earlier actions past perfect, habits simple past, and general truths present. Build a sequence. The question is not “Are all tenses the same?” but “Are the relationships coherent?”

8. Present perfect and finished time

The present perfect commonly links a past event to the present and avoids definite finished past-time expressions such as yesterday in standard usage. Thus I have seen him yesterday is usually edited to I saw him yesterday. But expressions such as today can be unfinished depending on context, so meaning still matters.

9. Past perfect overuse

Learners sometimes add had to every earlier past event. Past perfect is useful when the order needs marking; once the temporal sequence is established, simple past may resume. Editing should not inflate the verb phrase without purpose.

10. Future time in subordinate clauses

After time conjunctions such as when, after, before and as soon as, English often uses present tense for future reference: When she arrives, we will begin. Learners may incorrectly write when she will arrive. The distinction becomes a frequent editing target.

11. Modal chains

Modal + base form is a powerful scan. could went, may requires, must completed all fail because the lexical verb after the modal is not base form. Perfect modal constructions add have + past participle: might have missed. Teach the whole chain, not isolated forms.

12. Passive chains

A passive is built with a form of be plus past participle, sometimes inside longer chains: has been completed, will be reviewed. Students may edit the wrong auxiliary if they do not see the chain. Bracket it as one unit.

13. Adjective versus adverb

Do not reduce the rule to “add -ly.” Some adjectives and adverbs share forms: fast, hard, late. Some -ly words are adjectives: friendly, lively. The structural question is what the word modifies. A linking verb can take an adjective complement: The soup tastes good, not necessarily well unless describing the act of tasting.

14. Participial adjectives

Bored and boring, interested and interesting, confused and confusing encode different roles. The -ed form often describes the experiencer; the -ing form describes the cause. The students were boring by the lecture should normally be bored. Meaning selects the form.

15. Noun forms after determiners

If a slot follows the, a, this, possessives or adjectives and functions as the head of a noun phrase, test whether a noun form is required. The rapid of growth signals a mismatch. Sentence architecture helps more than suffix memorization.

16. Count versus noncount quantifiers

Many/few tend to pair with plural count nouns; much/little with noncount nouns. Fewer typically compares countable items; less noncount quantity, though everyday usage contains exceptions and fixed expressions. Follow the assessment standard without overstating the universality.

17. Article choice at first and later mention

A passage may introduce a researcher and later refer to the researcher. This is discourse grammar. An article error can only be solved by reading earlier context. Train learners to notice first mention versus established reference.

18. Generic reference

Plural and mass nouns can appear without articles when speaking generally: Libraries preserve knowledge; Exercise improves health. Over-inserting the is a common editing error. Ask whether the noun refers to a specific recoverable set or the class in general.

19. Prepositions after adjectives

Store adjective-preposition pairings as chunks: interested in, responsible for, similar to, aware of, capable of, dependent on. Use dictionaries to verify doubtful combinations during training.

20. Prepositions after verbs

Some verbs take no preposition before their object: discuss a problem, enter a room, approach a person. Others require one: depend on, belong to, apply for. The learner’s first language can influence these patterns, so error logs should store the complete frame.

21. Verb + gerund patterns

Common verbs followed by -ing forms include avoid, consider, enjoy, suggest and finish. The list is not enough; practise them in passages. She suggested to leave is a common error because many other verbs do take infinitives.

22. Verb + infinitive patterns

Common verbs followed by infinitives include decide, hope, plan, refuse, agree and promise. Some verbs allow both patterns with meaning changes, such as remember doing versus remember to do. Advanced editing needs semantic awareness.

23. Causative and perception patterns

After make in active causative use, standard English often uses object + base verb: made him leave. Allow uses object + infinitive: allowed him to leave. Perception verbs may take base or -ing with meaning differences. These frames are excellent advanced editing targets.

24. Pronoun case after prepositions

Prepositions typically take object-case pronouns: between you and me, for her and him. Hypercorrection creates between you and I. The fix follows function, not formality.

25. Reflexive pronouns

Myself is not a formal substitute for me or I. Reflexives usually refer back to the subject or add emphasis: I cut myself; I did it myself. Please contact myself is often an editing target in formal standard English.

26. Ambiguous reference

Not every ambiguous pronoun is a single-word grammar error, but advanced editing should train the eye. When two antecedents share compatible number and gender, repeating the noun may be the safest rewrite. In constrained single-word tasks, however, do not invent a correction the format cannot support.

27. Misplaced only

Only can change meaning depending on position: Only Iona solved the problem, Iona only solved the problem, Iona solved only the problem. Editing placement requires identifying what is being limited.

28. Dangling participles

Introductory -ing or -ed phrases usually need a nearby subject that can logically perform or experience the action. Built in 1920, tourists still visit the bridge makes tourists the thing built. Repair by making the bridge the subject of the main clause or restructuring.

29. Faulty comparisons

Compare like with like. The salary of engineers is higher than teachers compares salary to people. Use than that of teachers or rewrite. This is partly logic and partly syntax.

30. Double comparatives and superlatives

More better and most easiest double-mark degree in standard English. Choose the inflectional or periphrastic form appropriate to the adjective: better, easiest, more effective.

31. Parallel infinitives

Writers may omit repeated to after the first infinitive without losing parallelism: to read, write and speak is fine. The issue is not visible repetition but matched grammatical roles. Do not create a false positive by demanding to before every verb.

32. Paired conjunctions

With not only…but also, the elements following each half should be structurally comparable. She not only enjoys reading but also to write should be aligned, for example enjoys not only reading but also writing.

33. Subordination fragments

A clause beginning because, although, when or if can be complete only when attached appropriately to a main clause, unless used deliberately in informal or rhetorical contexts. Examination standards usually require complete sentence structures.

34. Comma splices

When punctuation is in scope, two independent clauses joined by a comma alone may require a full stop, semicolon, coordinating conjunction or subordination. Choose the repair according to the logical relationship.

35. Redundant conjunctions

Pairs such as Although…but and Because…so are often rejected in formal edited English when both elements mark the same relationship. Some language varieties permit patterns that others avoid. Train to the target standard.

36. Redundant comparatives

Expressions such as more superior can be redundant because superior already encodes comparison. Yet lexical usage is complex; verify doubtful cases rather than applying a simplistic “never modify” rule to every adjective.

37. Collocation disguised as grammar

Do a decision is structurally possible but nonstandard because English conventionally says make a decision. If collocation is within the task, fix it. If the task guarantees strictly grammatical errors, be cautious about treating every unusual word pairing as the target.

38. Register

Some forms are grammatical but informal: contractions, conversational particles, phrasal choices. An editing task may or may not test register. Read the rubric. Do not replace informal but grammatical language merely to sound academic unless the task requires formal register.

39. Punctuation around nonrestrictive clauses

Nonrestrictive relative clauses are typically set off by commas in standard written English: My aunt, who lives in Perth, is visiting. Restrictive clauses are not: The student who won the prize is absent. Meaning determines punctuation.

40. Final answer transfer

A learner can solve the grammar and still lose the mark by copying the wrong line number, misspelling the replacement or changing the wrong word. Treat answer transfer as part of execution. The final scan should verify not just the sentence but the recorded response.

Twenty additional mixed items with explanations

1. The purpose of the new rules are to reduce confusion.
2. A number of students is waiting outside.
3. The number of complaints have fallen this month.
4. She might has forgotten the appointment.
5. The repairs should be completed by Friday.
6. He has been work here since 2022.
7. The device operates automatic when the temperature rises.
8. We received many useful feedback from users.
9. The researchers are capable to repeat the experiment.
10. She insisted on checking the figures herself.
11. Neither Maren nor her classmates was ready.
12. Every one of the answers appear plausible.
13. The report, together with its appendices, were submitted yesterday.
14. There is several reasons for the delay.
15. By next month, the team will finish the first phase.
16. The manager suggested that the meeting was postponed.
17. Iona looks forward to analyse the new data.
18. The teacher made the students to rewrite the paragraph.
19. After reviewing the evidence, Leonie changed her answer.
20. The new system is more efficient than the old one.

Answers and reasons. 1 are → is; head purpose. 2 is → are; a number of means several. 3 have → has; the number is singular. 4 has → have; modal perfect chain might have forgotten. 5 correct. 6 work → working; present perfect progressive chain. 7 automatic → automatically; adverb modifies operates. 8 many → much or a great deal of under noncount feedback; some contemporary usage increasingly treats feedback flexibly, so follow assessment convention. 9 to repeat → of repeating. 10 correct. 11 in a proximity-agreement convention, was → were because the nearer subject classmates is plural. 12 appear → appears; head Every one. 13 were → was; parenthetical together with… does not make the subject plural. 14 is → are in formal agreement with several reasons. 15 can be correct: simple future can describe expected completion by a future time; future perfect will have finished would emphasize completed-before-deadline meaning but should not be forced without context. 16 if the intention is a recommendation for future action, many formal standards prefer subjunctive/base be postponed; check target convention. 17 analyse → analysing because to is a preposition in look forward to. 18 delete to; make + object + base verb. 19 correct. 20 correct.

Forty-item final mastery audit

Use this set without family labels. Mark each item correct or supply the minimum repair. Then classify your own errors before reading the key.

1. The quality of the submissions have improved.
2. The students in the final group have completed the task.
3. Each of the machines require a separate code.
4. The evidence suggests a different explanation.
5. Neither of the routes lead directly to the station.
6. By the time Leonie arrived, the lesson already started.
7. Maren has visited the exhibition yesterday.
8. Iona has been reading the report for an hour.
9. The results might indicates a change.
10. The documents were send by courier.
11. The experiment was remarkably success.
12. The device responded slow to the command.
13. She gave us several useful advice.
14. We need an information about the schedule.
15. The committee discussed the problem in detail.
16. The outcome depends of several factors.
17. He is responsible for maintaining the database.
18. The students avoided to repeat the error.
19. Maren decided revising the paragraph.
20. Iona looks forward to presenting her findings.
21. Between you and I, the answer seems doubtful.
22. The manager asked Leonie and I to stay.
23. After reading the message, the mistake became obvious.
24. Walking through the library, Iona noticed a new display.
25. The course aims to build confidence, improving accuracy and develop speed.
26. The new model is more cheaper than the previous one.
27. The rainfall here is higher than the coastal city.
28. The researcher who she designed the study presented the results.
29. The books that is missing have been replaced.
30. Although the road was closed, but traffic remained heavy.
31. Because the weather changed suddenly.
32. The alarm sounded, everyone left the building.
33. The team collected the samples and stored them immediately.
34. There are many possible explanations.
35. There is several issues to consider.
36. The number of visitors has increased steadily.
37. A number of visitors has complained about the queue.
38. The report was completed on time.
39. The committee has chosen a new representative.
40. Every answer should be checked before submission.

Mastery key. 1 have → has. 2 correct. 3 require → requires. 4 correct. 5 lead → leads under singular Neither convention. 6 already started → had already started. 7 has visited → visited with finished past yesterday. 8 correct. 9 indicates → indicate. 10 send → sent. 11 success → successful. 12 slow → slowly. 13 replace several useful advice with a count-compatible expression such as several useful pieces of advice. 14 remove an or use a piece of information. 15 correct. 16 of → on. 17 correct. 18 to repeat → repeating. 19 revising → to revise. 20 correct. 21 I → me. 22 I → me. 23 dangling modifier; make the reader the subject, e.g. After reading the message, Maren found the mistake obvious. 24 correct. 25 restore parallelism, e.g. to build confidence, improve accuracy and develop speed. 26 remove more. 27 compare rainfall with rainfall: higher than that in the coastal city. 28 delete she. 29 is → are. 30 remove but or restructure. 31 fragment; attach a main clause. 32 comma splice; use a full stop, semicolon or conjunction. 33 correct. 34 correct. 35 is → are. 36 correct. 37 has → have. 38 correct. 39 correct. 40 correct.

If you scored well but changed many correct items, your next priority is false-positive control. If you missed many errors but repaired nearly all that you noticed, train detection. If you noticed the errors but supplied incorrect forms, rebuild the grammar families. If untimed performance is strong but this set becomes unstable under a clock, train execution and recovery.

The point of the lane

A grammar editing question looks small because the answer may be one word. The cognitive task is not small. The learner must parse language, infer meaning, retrieve a grammatical relationship, discriminate error from acceptable variation, generate a minimal repair, verify it and manage time. That is why simple answer sheets can produce years of practice without reliable transfer.

This lane treats editing as an examination-performance system rather than a pile of corrections. The next step is to train the first bottleneck explicitly: detection before repair.

Keep correct language intact. Editing Accuracy — distinguish real errors from acceptable language and avoid unnecessary changes.

Federated performance loop: keep grammar detection, proof and repair inside SETC while the problem is genuinely English editing. If the same errors persist despite demonstrated rule knowledge, use Yishun Student Diagnostics to test whether the first weak link is retrieval, attention, transfer or task conditions. After repair, use Sengkang’s Assessment Evidence to interpret a fresh unseen editing sample; if the failure appears only under examination constraints, continue to How Examination Performance Works.

Owner return: this page applies grammar and sentence knowledge to examination editing. Use English Grammar and Sentence Control when the underlying relationship needs explanation; use How Feedback Changes Performance when a corrected error must change the learner’s next editing decision; and use How English Learning Transfers to test whether the repair survives a fresh item and later original writing. Return here for timed detection, justification and minimal correction under Secondary examination conditions.