Paragraph Completion Questions: Strategy and Worked Examples

Paragraph completion questions appear in the verbal sections of TCS NQT, Cocubes, and Infosys InfyTQ assessments, asking you to select one sentence from four options to fill a gap or complete a passage.

The format looks simple. In practice, roughly one in three students loses marks here by picking the option that mentions the right topic rather than the option that fits the paragraph’s logical arc. That gap in performance comes down to strategy, not vocabulary size.

What Paragraph Completion Items Actually Test

These questions do not test how many words you know. They test whether you can read a paragraph as a connected unit of meaning, not as a collection of independent sentences.

Specifically, they assess three things:

• Discourse coherence: whether the candidate sentence maintains the logical thread of the paragraph
• Register consistency: whether the candidate sentence matches the formal or informal tone of the surrounding text
• Referential accuracy: whether any pronouns or demonstratives in the candidate sentence (this, these, it, they) refer clearly to something already established in the passage

A student reviewing the TCS iON National Qualifier Test (NQT) verbal section guidance will find these three dimensions reflected in every sample item.

Three Types of Blank Positions

The blank can appear in three positions, and each position requires a slightly different approach.

Missing Last Sentence

The paragraph is complete except for the final sentence. The task is to identify the logical conclusion: a sentence that follows from everything stated above it.

Watch for options that introduce new topics at the end. A paragraph about ocean pollution should not conclude with a sentence about urban traffic.

Missing Middle Sentence

A sentence from the interior of the paragraph is removed. The correct option must connect to the sentence before it AND create a smooth transition to the sentence that follows.

This type is slightly harder because it carries two logical constraints. The blank must satisfy the sentence above (incoming coherence) and the sentence below (outgoing coherence).

Missing First Sentence

The introductory sentence is removed. The remaining paragraph usually contains a strong cue: “this development,” “these measures,” “that approach.” The opening sentence must introduce the entity, action, or idea that those reference words point back to.

If the second sentence begins with “This surge in computational speed,” the missing first sentence almost certainly describes what caused that surge.

A Systematic Strategy for Any Blank Position

Five steps, in order:

• Step 1 — Identify the topic: Read the complete passage (gap included) and state the topic in one phrase. “Impact of microplastics on marine food chains.” “India’s renewable energy growth trajectory.”
• Step 2 — Identify the tone: Is this a scientific description? A policy argument? A personal narrative? A news report? The correct answer will match this tone. Eliminate any option that shifts register.
• Step 3 — Map the logical flow: Is the paragraph presenting a cause-and-effect chain? A chronological sequence? An argument with supporting evidence? Identify the direction, then check which option continues it.
• Step 4 — Eliminate aggressively: Any option that (a) introduces a completely new topic, (b) contradicts the paragraph’s main claim, or (c) uses vocabulary inconsistent with the passage register can be dropped immediately. Most items have at least two clear eliminations.
• Step 5 — Test the survivors: Read each surviving option into the blank and read the full paragraph in your head. Check pronoun referents. Check logical connectors. The correct sentence should feel inevitable, not merely possible.

This five-step approach handles the majority of paragraph completion items in any campus placement test.

Five Worked Examples with Full Reasoning

Each example below lays out all four options, derives the correct answer from discourse-level cues, and verifies the result. These items are original; the answer derivation is done from first principles.

Example 1: Missing Last Sentence

• Passage: Microplastics have been detected in ocean sediments, marine organisms, and even in the tissues of fish consumed by humans. Researchers have tracked their spread from coastal areas to the deep ocean floor. Scientists have documented how these particles pass through the food chain, accumulating in organisms at each trophic level. _________
• (a) The economic impact of plastic packaging on retail businesses has grown steadily since the 1990s.
• (b) This bioaccumulation pattern means that larger predators, including humans, end up with higher concentrations of microplastics in their systems.
• (c) Plastic waste also contributes to urban pollution, particularly in densely populated cities.
• (d) Several governments have introduced legislation to regulate the sale of single-use plastics in supermarkets.
• Topic: Microplastic contamination in the marine food chain.
• Tone: Scientific, third-person, factual.
• Flow: Detection → spread → food chain accumulation at each trophic level.
• Elimination: (a) shifts to retail economics — off-topic. (c) shifts to urban land-based pollution — wrong domain. (d) introduces policy response — breaks the scientific description register.
• Survivor: (b) directly follows from “accumulating at each trophic level”: the logical conclusion is that apex predators (including humans) carry the highest concentrations.
• Answer: (b)

Example 2: Missing Middle Sentence

• Passage: Online learning platforms have made quality education accessible to students in remote areas across India. _________. However, the lack of immediate feedback and classroom interaction remains a limitation for learners in self-paced formats.
• (a) The cost of printed textbooks has increased in tier-2 cities over the past five years.
• (b) A student in Nagpur or Patna can now access lectures from leading institutions without relocating to a metro.
• (c) Several universities have recently upgraded their physical campuses with modern laboratory facilities.
• (d) Governments have invested in rural road infrastructure to improve physical access to colleges.
• Topic: Online learning and its accessibility benefit.
• Tone: Balanced, analytical.
• Flow: Accessibility benefit (first sentence) → [blank] → concession of a drawback (third sentence).
• Elimination: (a) textbook costs — unrelated to online learning. (c) physical campus upgrades — contradicts the online theme. (d) road infrastructure — physical access, not online.
• Survivor: (b) provides a concrete India-specific example of the accessibility benefit (Nagpur, Patna), bridging the general first-sentence claim to the contrasting limitation in the third sentence.
• Answer: (b)

Example 3: Missing First Sentence

• Passage: _________. This surge in computational speed allowed large datasets to be processed in hours rather than months. Combined with advances in neural network design, it enabled modern applications like image recognition and language translation.
• (a) The internet became commercially available in the 1990s, transforming global communication patterns.
• (b) The development of graphical processing units (GPUs) in the early 2000s dramatically increased the speed of parallel computations.
• (c) Machine learning algorithms were first proposed by academic researchers working in university laboratories.
• (d) Cloud computing services have allowed businesses of all sizes to reduce their infrastructure costs.
• Topic: Cause of the computational speed surge that enabled AI applications.
• Tone: Technical-explanatory, factual.
• Flow: [cause] → speed surge → neural network design → AI applications.
• Key cue: The second sentence opens with “This surge in computational speed.” The opening sentence must introduce what caused this surge.
• Elimination: (a) the internet and global communication — no referent for “this surge.” (c) ML algorithm proposals — does not explain a speed surge. (d) cloud computing for businesses — different cause, different era.
• Survivor: (b) GPUs → parallel computation speed = the direct antecedent of “this surge.” The causal chain is complete and the referent is unambiguous.
• Answer: (b)

Example 4: Missing Middle Sentence

• Passage: The rise of social media has changed how political campaigns engage with voters. Candidates can broadcast messages directly to millions without relying on traditional media intermediaries. _________. Political analysts note that digital engagement consistently outperforms television advertising among first-time voters.
• (a) Television advertising remains the dominant medium in rural constituencies.
• (b) This direct channel has also created new challenges, as false information spreads rapidly through the same platforms.
• (c) Social media companies generate advertising revenue during national election cycles.
• (d) Campaign finance laws in many democracies impose limits on total advertising expenditure.
• Topic: Social media’s role in election campaigns.
• Tone: Analytical, neutral, policy-adjacent.
• Flow: Direct communication benefit → [blank] → engagement data favouring digital.
• Elimination: (a) shifts to TV in rural areas — contradicts the social media focus. (c) shifts to company revenue — different actor (platforms, not campaigns). (d) introduces campaign finance — new topic.
• Survivor: (b) extends the social media discussion by introducing a countervailing risk, which makes the final sentence’s engagement data more substantive in context.
• Answer: (b)

Example 5: Missing Middle Sentence

• Passage: India’s renewable energy capacity has grown sharply over the past decade. Solar installations crossed 70 GW of installed capacity in 2024, up from less than 3 GW in 2014. _________. The government has set a target of 500 GW of renewable energy by 2030.
• (a) Coal remains the dominant source of electricity generation in India, accounting for over 50 per cent of output.
• (b) Wind energy has also expanded steadily, with installed capacity exceeding 45 GW, concentrated in Tamil Nadu, Gujarat, and Rajasthan.
• (c) Hydroelectric power plants were among the first forms of renewable energy deployed at scale in India.
• (d) Several large infrastructure projects in India have faced delays due to land acquisition challenges.
• Topic: India’s renewable energy growth trajectory.
• Tone: Data-oriented, forward-looking.
• Flow: General growth claim → solar data (2014 → 2024) → [blank] → 2030 future target.
• Elimination: (a) coal dominance — contradicts the renewable growth theme. (c) historical hydroelectric — breaks the forward chronology (the paragraph moves from 2014 to 2024 to 2030; going back to founding-era hydroelectric is out of sequence). (d) infrastructure delays — tangential.
• Survivor: (b) extends the renewable growth narrative with wind energy data, completing the picture before the paragraph moves to the 2030 target.
• Answer: (b)

Common Traps That Eliminate the Wrong Option First

Three patterns account for most wrong answers on paragraph completion items.

The related-but-wrong option. This option mentions the correct topic but shifts to a sub-topic the paragraph never addresses. In Example 1, option (d) mentions plastic regulation, which is related to microplastics but introduces a policy dimension that the scientific description never established. Topic proximity is not enough.

The over-specific distractor. This option provides a highly specific fact or statistic that sounds authoritative but does not logically connect to the sentence before or after the blank. A sentence with a precise number attached to it can feel conclusive, but the number must actually advance the paragraph’s argument.

The register-shift option. This option matches the topic but changes the level of formality, or shifts from objective description to opinion, or from policy argument to personal anecdote. Tonal inconsistency is a reliable eliminator and it works quickly.

For verbal tests at companies with analytical, multi-step hiring processes, like the D.E. Shaw verbal section or the Sopra Steria verbal test, paragraph completion items appear alongside jumble and sentence-correction questions. The same topic-tone-flow framework that works for paragraph completion transfers directly to jumble questions, where logical sequencing of sentences is the core task.

The discourse analysis skill that paragraph completion tests is also close to what language models do when they predict the next token in a sequence. If you want to see how that works in practice, TinkerLLM lets you run hands-on experiments with text generation and coherence for ₹299 a month. It’s a short leap from understanding how a paragraph hangs together logically to understanding why an AI completes it the way it does.