Sentence Correction Basics: Grammar for Placement Tests

Sentence correction appears in every major campus placement test in India: AMCAT, TCS NQT, Cocubes, and company-specific online assessments all include it as a graded section.

The format is consistent. You get a sentence, sometimes with one underlined portion and sometimes with four labelled segments. You choose which version corrects the error, or which labelled part is wrong. The skill being tested is structural: can you identify a grammatical error quickly, without relying on feel alone?

Students who know the error types and their fix rules answer these questions in 30 to 40 seconds each. Students who rely on “this sounds right” stall on the tricky variants designed to make the wrong option sound natural. This article covers the most-tested rule (subject-verb agreement), the seven patterns under it, and how to connect the basics to the broader sentence correction error taxonomy.

How Sentence Correction Questions Are Structured

Two formats appear across placement tests:

• Full-sentence format: A complete sentence is presented. One or more answer options show different versions of the same sentence. Choose the version with no grammatical error.
• Underline or label format: Part of the sentence is underlined or labelled A, B, C, D. You choose which label contains the error, or pick “No error” if the sentence is correct.

The second format is more common in AMCAT English and Cocubes verbal sections. TCS NQT tends to use the first format more heavily. In both cases, the approach is the same: identify the error type, apply the rule, pick the option that fixes it.

Error types are not distributed evenly across tests. Subject-verb agreement accounts for a disproportionate share of sentence correction questions, followed by tense consistency, pronoun reference, and modifier placement. Starting with subject-verb agreement is the most effective starting point for any student beginning verbal prep.

Subject-Verb Agreement: The Core Rule

The base rule: the verb must match its subject in number and person. A singular subject takes a singular verb; a plural subject takes a plural verb.

Purdue OWL’s subject-verb agreement guide is the standard reference for these patterns. Seven patterns cover the full range of variants tested in placement assessments.

Pattern 1: Simple Singular and Plural

• Singular subject takes a singular verb.
• Plural subject takes a plural verb.
• Correct: She reads every placement paper from the last three years.
• Correct: They read every placement paper from the last three years.

Pattern 2: Compound Subjects Joined by “And”

• Two subjects joined by “and” form a plural subject and take a plural verb.
• Correct: Ravi and Priya are preparing for TCS NQT together.
• Exception: If the two nouns refer to the same entity or concept, use a singular verb.
• Correct: Bread and butter is a common breakfast item. (One concept.)

Pattern 3: Compound Subjects Joined by “Or” or “Nor” (Proximity Rule)

• The verb agrees with the subject closest to it.
• Correct: Neither the manager nor the trainees were informed.
• Correct: Neither the trainees nor the manager was informed.
• This pattern inverts many students’ intuition, which is why it appears frequently on tests.

Pattern 4: Collective Nouns

• Collective nouns (team, jury, committee, class, panel) are singular when the group acts as one unit.
• Correct: The panel has finalised the shortlist. (Acting as one unit.)
• Correct: The panel are divided on the final candidate. (Members acting individually.)
• Most placement test sentences use the group-as-unit reading, so default to singular unless the sentence signals individual action explicitly.

Pattern 5: Indefinite Pronouns

• Singular indefinite pronouns (each, everyone, someone, nobody, anyone, either) take singular verbs.
• Correct: Everyone is required to submit the form before the deadline.
• Correct: Each of the candidates has received a confirmation email.
• Plural indefinite pronouns (few, many, several, both) take plural verbs.
• Correct: Few have completed the online assessment on time.
• Context-dependent pronouns (some, any, none, all) agree with the noun they refer to.
• Correct: Some of the data is missing. (Uncountable noun; singular verb.)
• Correct: Some of the files are missing. (Countable plural noun; plural verb.)

Pattern 6: Existential Constructions

• In sentences beginning with “there” or “here”, the verb agrees with the subject that follows, not with “there” or “here”.
• Correct: There is one grammar rule every test-taker must know.
• Correct: There are three error types covered in this section.

Pattern 7: Titles and Names of Organisations

• Titles of books, films, courses, and organisations take singular verbs, even when the name looks plural.
• Correct: “Great Expectations” is required reading in that module.
• Correct: Infosys has released its campus hiring calendar for this year.

The Prepositional-Phrase Trap

The most common exploit used by question-setters is the intervening prepositional phrase. A phrase like “of the students” or “between the candidates” placed between the subject and verb does not change the subject’s number. Students who read quickly trip over this every time.

The fix: strip the prepositional phrase and read subject → verb directly.

• Incorrect: The list of items are on the table.

• Correct: The list of items is on the table. Strip “of items”; the subject is “list”, which is singular.

• Incorrect: Each of the students have a textbook.

• Correct: Each of the students has a textbook. Strip “of the students”; the subject is “each”, which is always singular.

• Incorrect: The group of engineers are ready to present.

• Correct: The group of engineers is ready to present. Strip “of engineers”; the subject is “group”, which takes a singular verb as a collective noun acting as a unit.

This stripping technique works on every variant of Pattern 1 through Pattern 7. It is the single fastest way to resolve subject-verb agreement questions under time pressure.

Other Error Types and Where to Go Next

Subject-verb agreement is the starting point. The full sentence correction error taxonomy covers six more types that appear in placement tests: pronoun reference, tense consistency, misplaced modifiers, parallelism, comparison errors, and preposition or article usage.

Two of those have dedicated deep dives in this series:

• Misplaced modifiers in sentence correction covers the four modifier types (misplaced, dangling, squinting, and limiting) that generate their own cluster of questions on AMCAT and TCS NQT verbal sections.
• Verb time sequence in sentence correction covers the six tense-agreement patterns tested, including past-perfect sequencing, present-perfect versus simple-past, and reported-speech backshift.

Cambridge Grammar’s reference on collective nouns is a concise check when working through collective-noun questions in unfamiliar sentence constructions.

Building Sentence Correction Accuracy

The two-pass method is the tactical approach for test conditions:

• Pass 1: Read the sentence for meaning. What is the sentence trying to say? Identify the main subject and the main verb.
• Pass 2: Apply the error-type checklist in priority order: subject-verb agreement first, then pronoun reference, then tense, then modifier placement, then parallelism. Stop as soon as a mismatch appears.

Most questions contain exactly one error. The two-pass approach is faster than trying to absorb the full sentence at once, and more reliable than feel-based reading on the variants designed to mislead.

For the drill plan, the verbal ability preparation guide lays out a 4-week cycle that slots sentence correction into a broader verbal schedule. The general sequence for sentence correction specifically:

• Week 1: Subject-verb agreement — all 7 patterns, 20 questions per day, timed.
• Week 2: Add pronoun reference and tense consistency; keep subject-verb agreement in rotation.
• Week 3: Add modifier placement and parallelism.
• Week 4: Mixed-error timed sets; 30 to 40 seconds per question.

Source material: AMCAT sample papers, previous-year TCS NQT verbal questions, and Cocubes verbal practice sets. All three draw from the same error-type pool, so cross-practice transfers cleanly.

Pattern recognition built through this drilling applies beyond test prep. Evaluating AI-generated text for structural errors (subject-verb mismatches, modifier attachment, tense inconsistencies) runs on exactly the same instinct. TinkerLLM at ₹299 is a live environment where that instinct meets real LLM outputs; sentence correction accuracy becomes a practical asset there, not just an exam score.