Critical Reasoning for Placement Tests: A Beginner's Guide

Critical reasoning questions test one thing: whether you can separate a stated fact from the conclusion drawn from it.

That skill appears across most major placement tests: TCS NQT, AMCAT, CAT, eLitmus, and the analytical reasoning sections used by consulting firms. The questions share a common structure and a small set of repeating question types. Learn the types and the gap they each probe, and the questions stop feeling arbitrary.

What Critical Reasoning Tests

Every CR question gives you a short passage of 3 to 6 sentences built around an argument. An argument has two parts:

• Premise: a stated fact, observation, or piece of evidence.
• Conclusion: a claim the author draws from that premise.

Pure logical reasoning (syllogisms, Venn diagrams) tests whether a formal inference rule applies correctly. Reading comprehension tests whether you understood what a passage said. Critical reasoning sits between them: the passage may contain a logical gap, and the question asks you to probe that gap.

The three things CR questions probe:

• Gap in reasoning: what the author assumed but never stated (assumption questions)
• What follows necessarily: what must be true given the stated premises (inference questions)
• How the argument can be affected: what additional facts would support or undermine the conclusion (strengthen and weaken questions)

This gap-analysis approach is also tested in data analytics and management consulting selection rounds. The Mu Sigma selection process includes analytical reasoning sections that apply similar argument-evaluation skills.

Six Question Types, Identified by Stem Keywords

Every CR question on TCS NQT, AMCAT, CAT, or eLitmus belongs to one of six types. Identifying the type from the question stem is half the work.

Type	What it asks	Common stem keywords
Assumption	Which unstated fact must be true for the argument to hold?	“assumes”, “takes for granted”, “depends on”
Strengthen	Which answer most supports the conclusion?	“strengthens”, “supports”, “most helps”
Weaken	Which answer most undermines the conclusion?	“weakens”, “undermines”, “calls into question”
Inference	What can be concluded from the passage?	“can be inferred”, “must be true”, “best supported”
Paradox	What explains the apparent contradiction?	“explains”, “resolve the discrepancy”, “reconcile”
Parallel reasoning	Which argument has the same logical structure?	“most similar in reasoning”, “parallel argument”

The two types students most often confuse are assumption and inference. The memory hook: assumptions go into the argument (the missing premise the author is relying on), while inferences come out of the argument (conclusions that must follow from what is stated).

Strengthen and weaken questions share the same structure. The only difference is direction. Both require finding the gap between premise and conclusion first.

The Three-Step Approach

The same three steps apply to every CR question type.

Step 1: Read for Structure

Read the passage once, marking the conclusion. Look for signal words: “therefore”, “thus”, “hence”, “so”, “this suggests that”, “as a result”. Everything before the conclusion signal is a premise. If there is no explicit signal word, the last sentence is usually the conclusion.

Step 2: Identify the Question Type

Read the question stem and match the keyword to the table above. This step is the one most beginners skip. Without naming the type, every answer choice looks equally plausible.

Step 3: Apply the Type-Specific Action

Pre-think your answer before reading the choices:

• For assumption questions: ask what the author needs to be true but never actually said. The correct choice is the one that, if removed, makes the argument collapse.
• For strengthen and weaken questions: find the gap between the premise and the conclusion first. Then ask whether the choice closes the gap (strengthen) or widens it (weaken).
• For inference questions: accept only what the passage requires to be true. If a choice goes one step beyond what is stated, it is not a valid inference.
• For paradox questions: look for a third fact that makes both contradictory parts true at the same time.
• For parallel reasoning questions: strip the specific content and write the abstract structure, then match that structure across answer choices.

Four Worked Examples

Working through examples reveals the approach faster than re-reading definitions. Each is solved step by step.

Assumption

• Passage: “The government should ban advertisements during children’s television programs. Research shows that children aged 5 to 10 cannot distinguish between a program and an advertisement.”
• Question: “The argument above assumes which of the following?”
• Step 1: Premise: children aged 5 to 10 cannot distinguish ads from programs. Conclusion: the government should ban ads during children’s TV.
• Step 2: “assumes” in the stem signals an Assumption question.
• Step 3: Pre-think the gap. The premise says children cannot tell ads apart from programs. The conclusion is to ban ads. The unstated bridge: that this confusion causes some harm justifying a ban. The author never states there is harm from the confusion.
• Answer: “Television advertising influences children’s behaviour more effectively when they cannot identify the content as advertising.”
• Why this is the assumption: Remove this premise and the argument collapses. If the confusion causes no identifiable harm, there is no basis for the ban.

Weaken

• Passage: “A new metro line connecting the city center to the airport reduced taxi bookings by 30% in its first month. The city government concluded that the metro will solve the city’s traffic congestion problem.”
• Question: “Which of the following, if true, most weakens the government’s conclusion?”
• Step 1: Premise: metro reduced taxi bookings by 30%. Conclusion: metro will solve traffic congestion.
• Step 2: “most weakens” signals a Weaken question.
• Step 3: The argument assumes taxi bookings are a primary congestion driver and that fewer taxis means less congestion. A weakener shows the conclusion does not follow from the premise.
• Answer: “Private car ownership in the city increased by 40% in the same month the metro opened, as commuters used the metro for one leg of the trip while adding personal vehicles for other trips.”
• Why this weakens: Car ownership growth offset the congestion relief from fewer taxis. The metro’s benefit is neutralised, so the conclusion that it will “solve” congestion is not supported.

Inference

• Passage: “No engineers in the placement cohort who passed the verbal aptitude cut-off failed the final selection round. Aditya is an engineer in the placement cohort who passed the verbal aptitude cut-off.”
• Question: “Which of the following must be true?”
• Step 1: Two premises. All verbal-pass engineers passed final selection; Aditya is a verbal-pass engineer.
• Step 2: “must be true” signals an Inference question.
• Step 3: Apply the structure: All A are B; Aditya is A; therefore Aditya is B. This conclusion follows necessarily.
• Answer: “Aditya passed the final selection round.”
• Why this is the only valid inference: Any choice claiming “Aditya is the strongest candidate” or “all similar engineers will always succeed” goes beyond what the premises state. The inference question requires exactly what must be true, not what might be true.

Paradox

• Passage: “City A introduced strict emissions controls on all factories last year. Despite this, air pollution levels in City A rose by 12% compared to the previous year.”
• Question: “Which of the following, if true, best explains the apparent discrepancy?”
• Step 1: The contradiction: factory controls should reduce pollution, but pollution increased.
• Step 2: “explains the apparent discrepancy” signals a Paradox question.
• Step 3: Look for a third factor that makes both facts true simultaneously. The controls worked on factories, yet pollution rose, so something else must have added more pollution than the controls removed.
• Answer: “A new expressway opened in City A last year, doubling the daily volume of private vehicle traffic.”
• Why this resolves the paradox: Factory controls cut one source; vehicle exhaust added a larger one. Both the effectiveness of the controls and the rise in pollution can be true at the same time.

Building Speed and Accuracy

The single biggest time drain on CR questions is re-reading the passage because you did not mark the conclusion on the first pass. Use a consistent annotation habit: circle or underline the conclusion word (“therefore”, “hence”, “so”) the moment you see it.

A second time-saver: eliminate answer choices that change the topic. In assumption and strengthen questions, a choice that introduces a subject not present in the passage is almost never correct. The correct answer stays within the argument’s scope.

IndiaBix’s verbal reasoning practice section organises CR questions by type, which is the most effective drilling format: 10 questions of one type at a time rather than a randomised mix. AMCAT’s official preparation page lists the current syllabus and sample questions by module, including the verbal and analytical sections where CR passages appear.

The broader placement preparation guide covers how CR fits into the full selection process alongside written tests, group discussion, and interview rounds.

The argument-parsing discipline built across the four question types above transfers directly to evaluating LLM outputs. A model answer has premises and inferences too, and the habit of asking which is which is worth developing early. TinkerLLM (₹299 at tinkerllm.com) applies that same premise-to-inference check on live model responses rather than on passage-based placement questions.