ProgrammesCrucibleScholarshipBlogDiagnostic TestApplyStudent Login
English · Micro-Test

Reading Comprehension: Inference Questions

Inference questions ask what a passage implies without stating outright, and they trip up more CLAT aspirants than any other RC sub-type.

10 questions · 5 minutes · instant scoring

What this topic actually tests

An inference question does not reward what you already know about the world; it rewards what the author's words logically permit you to conclude and nothing more. CLAT passages for this sub-type are usually argumentative or editorial in tone, running to roughly 400-450 words, and they tend to build a claim through qualifiers, contrasts, and cause-effect links rather than flat assertions. The correct answer is the one that follows necessarily from the stated premises, even if it is never spelled out in so many words. The method is to treat the passage like a set of premises in a logical argument: identify the author's central claim, note every qualifying word ("often", "primarily", "in most cases"), and ask which answer choice is a natural, minimal extension of those premises rather than a leap beyond them. A useful check is to ask, "Could the author deny this statement without contradicting anything already written?" If yes, it is not a valid inference. Consider a mini-example: a passage states that a regulator relaxed compliance rules for fintech startups "to allow experimentation before a full market launch," and that several startups used this window to test products with real customers under supervision. A valid inference is that the regulator prioritised controlled real-world testing over blanket restriction — this follows directly. An invalid inference would be that the regulator generally favours deregulation across all sectors, since the passage only discusses one mechanism in one sector. Strong inference answers are narrow, textually anchored, and often restate the passage's logic in different words; weak ones import outside knowledge, exaggerate scope, or reverse a causal relationship the author never asserted.

The common trap on this topic

The single most common error is choosing an option because it is true in the real world or matches the aspirant's prior knowledge, even though the passage never actually supports it — examiners deliberately plant such "outside-knowledge" distractors because they feel intuitively correct. A close second is picking an answer that is too extreme: the passage might say a practice is "sometimes risky," and the trap option converts this into "always dangerous" or "should be banned," which overreaches what the text licenses. A third trap is the reversed-causation option, where the passage says A contributes to B, and the distractor claims B causes A, or that removing A would automatically produce the opposite of B. Aspirants under time pressure also fall for options that use several exact words from the passage but recombine them into a claim the author did not make — this "familiar vocabulary, false logic" trap is common in CLAT specifically because passages are dense and skimmed quickly. The fix is to physically point to the sentence(s) that justify each word of the chosen answer before selecting it.

Take the micro-test

Time Left
05:00
Answered
0 / 10
Judicial review, the power of courts to strike down laws that violate the Constitution, is often defended as a check on majoritarian excess. Yet the doctrine carries an unavoidable tension: unelected judges are empowered to override decisions made by elected representatives. Defenders argue this tension is a feature, not a flaw, because constitutional rights exist precisely to survive shifts in popular opinion. Critics counter that judicial review can itself become a vehicle for entrenching the preferences of a narrow legal elite, particularly when courts interpret vague constitutional text expansively. Both camps agree, however, that the legitimacy of judicial review depends on courts exercising the power sparingly and with reasoned restraint, rather than as a routine substitute for legislative policy-making.
Q1.

It can be inferred from the passage that judicial review loses legitimacy, according to both camps, when:

Judicial review, the power of courts to strike down laws that violate the Constitution, is often defended as a check on majoritarian excess. Yet the doctrine carries an unavoidable tension: unelected judges are empowered to override decisions made by elected representatives. Defenders argue this tension is a feature, not a flaw, because constitutional rights exist precisely to survive shifts in popular opinion. Critics counter that judicial review can itself become a vehicle for entrenching the preferences of a narrow legal elite, particularly when courts interpret vague constitutional text expansively. Both camps agree, however, that the legitimacy of judicial review depends on courts exercising the power sparingly and with reasoned restraint, rather than as a routine substitute for legislative policy-making.
Q2.

The defenders of judicial review mentioned in the passage would most likely agree that:

The city is not simply a collection of buildings but an accumulation of choices about whose convenience matters. A flyover routed around an affluent neighbourhood while cutting through a working-class settlement is not a neutral engineering decision; it is a statement of whose commute time was worth preserving. Planners rarely announce such priorities openly. Instead, they are expressed through technical language — "least-cost alignment," "minimal land acquisition" — that obscures the human calculus underneath. Reading a city's infrastructure, then, requires treating technical justifications with a degree of suspicion, not because engineers are dishonest, but because the vocabulary of efficiency can quietly launder decisions that are, at bottom, distributive.
Q3.

The passage implies that technical planning language such as "least-cost alignment" is problematic mainly because it:

The city is not simply a collection of buildings but an accumulation of choices about whose convenience matters. A flyover routed around an affluent neighbourhood while cutting through a working-class settlement is not a neutral engineering decision; it is a statement of whose commute time was worth preserving. Planners rarely announce such priorities openly. Instead, they are expressed through technical language — "least-cost alignment," "minimal land acquisition" — that obscures the human calculus underneath. Reading a city's infrastructure, then, requires treating technical justifications with a degree of suspicion, not because engineers are dishonest, but because the vocabulary of efficiency can quietly launder decisions that are, at bottom, distributive.
Q4.

Which of the following can be reasonably inferred about the author's view of engineers?

Artificial intelligence systems trained on historical hiring data have, in several documented cases, learned to penalise resumes associated with women or minority applicants, simply because past hiring at the company skewed toward other groups. The algorithm is not "biased" in the sense of holding an opinion; it is faithfully reproducing a pattern in its training data. This distinction matters for remedy: fixing the code alone does not fix the problem if the underlying data continues to encode historical inequity. A model retrained on the same skewed dataset, however cleverly re-engineered, will tend to rediscover the same pattern it was told to avoid, unless the data itself is deliberately rebalanced or the historical proxy variables are identified and removed.
Q5.

The passage suggests that simply rewriting an AI hiring algorithm's code, without addressing its training data, would most likely result in:

Artificial intelligence systems trained on historical hiring data have, in several documented cases, learned to penalise resumes associated with women or minority applicants, simply because past hiring at the company skewed toward other groups. The algorithm is not "biased" in the sense of holding an opinion; it is faithfully reproducing a pattern in its training data. This distinction matters for remedy: fixing the code alone does not fix the problem if the underlying data continues to encode historical inequity. A model retrained on the same skewed dataset, however cleverly re-engineered, will tend to rediscover the same pattern it was told to avoid, unless the data itself is deliberately rebalanced or the historical proxy variables are identified and removed.
Q6.

It can be inferred that the author distinguishes between an algorithm being "biased" and an algorithm being:

Judicial review, the power of courts to strike down laws that violate the Constitution, is often defended as a check on majoritarian excess. Yet the doctrine carries an unavoidable tension: unelected judges are empowered to override decisions made by elected representatives. Defenders argue this tension is a feature, not a flaw, because constitutional rights exist precisely to survive shifts in popular opinion. Critics counter that judicial review can itself become a vehicle for entrenching the preferences of a narrow legal elite, particularly when courts interpret vague constitutional text expansively.
Q7.

Based on the passage, critics of judicial review would most likely support:

The city is not simply a collection of buildings but an accumulation of choices about whose convenience matters. A flyover routed around an affluent neighbourhood while cutting through a working-class settlement is not a neutral engineering decision; it is a statement of whose commute time was worth preserving.
Q8.

The example of the flyover is used in the passage primarily to:

Artificial intelligence systems trained on historical hiring data have, in several documented cases, learned to penalise resumes associated with women or minority applicants, simply because past hiring at the company skewed toward other groups.
Q9.

From the passage, one can infer that the historical hiring data referenced most likely:

Judicial review, the power of courts to strike down laws that violate the Constitution, is often defended as a check on majoritarian excess. Yet the doctrine carries an unavoidable tension: unelected judges are empowered to override decisions made by elected representatives.
Q10.

Which of the following, if true, would most weaken the "tension" the passage describes between judicial review and democracy?

FAQ

How is an inference question different from a direct/literal question in CLAT RC?

A literal question is answered by a sentence stated almost verbatim in the passage. An inference question requires you to combine two or more stated ideas, or read between the lines of a qualifier, to reach a conclusion the author implies but never writes directly.

Is it ever correct to use outside knowledge to answer an inference question?

No. CLAT inference questions are self-contained — the correct answer must be derivable from the passage alone. If an option is factually true in general but unsupported by the specific passage, it is a trap, not the answer.

How much time should I spend per inference question in the actual exam?

Roughly 60-75 seconds once you have already read the passage, since you should be scanning for the relevant lines rather than re-reading the whole passage each time.

Keep practising

Ready to see the full picture?

This micro-test covers one topic. The full CLAT diagnostic covers all five sections in 30 minutes and gives you a section-wise breakdown of where you actually stand.

Take the Full Diagnostic →