CSAT Paper-II: Comprehension, Decision Making and Interpersonal Skills

CSAT is the paper that ends the most careers by accident. It is qualifying only at 33%, so no mark above the cut-off counts towards your rank — and yet failing it makes the GS Paper-I score irrelevant, however good it was. The paper has 80 questions for 200 marks in two hours, with negative marking of one-third on wrong answers, so you need roughly 27 correct answers out of 80 to be safe. The numeracy, data interpretation and logical reasoning halves are covered by the shared Mental Ability chapters. This chapter covers what those do not: comprehension, which is the largest single block in the paper, decision making and problem solving, which is the only section with no negative marking and a scoring convention unlike anything else in the exam, and interpersonal and communication skills. Treat the paper as a 33% target rather than a maximising exercise, and the strategy below becomes obvious.

The Arithmetic of Qualifying — Work Out Your Target Before You Start

CSAT at a glance — and what the numbers imply
FeatureValueWhat it implies for strategy
Questions and marks80 questions, 200 marks, 120 minutes2.5 marks each, 90 seconds each — but comprehension needs more and can afford it because reasoning needs less
Qualifying mark33% = 66 marksRoughly 27 correct answers with no wrong ones; about 33–35 correct if you also attempt some and miss a few
Negative markingOne-third of 2.5 ≈ 0.83 per wrong answerThree wrong answers cancel one correct — so blind guessing is mathematically negative
Decision-making sectionConventionally no negative marking, and some options earn partial creditNever leave a decision-making question blank — there is no downside
🎯 Why a qualifying paper deserves real preparation anyway
Because the failure mode is catastrophic and asymmetric. A candidate who scores 120 in GS Paper-I and 60 in CSAT is out; a candidate who scores 90 and 70 goes through. Every year, otherwise strong candidates fail on a paper worth no rank at all — usually because they had not practised comprehension under time pressure, or because they guessed their way into a negative score. Two to three full papers under strict timing, plus deliberate practice on whichever half is weaker, is normally the whole requirement. Do not over-invest either: marks above 66 are literally worthless.

Comprehension — the Largest Block, and Entirely Trainable

CSAT comprehension passages are typically short — one or two paragraphs — drawn from ethics, ecology, governance, economics, philosophy and social issues, and followed by one to three questions. Crucially, the answer is always inside the passage. This is not a general-knowledge test, and the single most common cause of error is importing your own knowledge or opinion into an option that the passage does not support. The instruction "on the basis of the passage alone" is meant literally.

  1. Read the passage first, not the questions. With passages this short, reading the questions first makes you scan for keywords and miss the argument, and CSAT questions are usually about the argument rather than a detail.
  2. Identify the author’s central claim in your own words before looking at the options. If you cannot state it in one sentence, read once more — the second reading of a short passage costs less than eliminating four options blindly.
  3. Then match, and eliminate by defect. Almost all wrong options fail in one of five ways: too extreme (uses "always", "never", "only"); too narrow, restating one detail as if it were the main point; too broad, generalising beyond what the passage licenses; outside the passage, true in the world but unsupported here; or reversed, stating the opposite of the author’s position. Naming the defect is faster than re-reading.
  4. Watch the qualifier words in both passage and options. "Some", "most", "may", "tends to" and "suggests" are load-bearing. A passage that says a policy "may reduce" inequality does not support an option asserting that it "eliminates" inequality — and that single substitution is one of the most frequently used traps in the paper.
⚠️ The "most logical and rational inference" question type
This phrasing appears constantly and has a specific meaning. You are not being asked what is true, nor what you agree with, nor even what follows possibly — you are being asked which option follows necessarily and minimally from the passage. So prefer the modest inference over the ambitious one. If option A says "governments should therefore ban the practice" and option B says "the practice has costs that its supporters understate", B is almost always the answer, because a passage describing a problem rarely licenses a specific policy prescription.

Decision Making and Problem Solving — the Only Section With No Downside

These questions place you in an administrative or workplace situation and offer four courses of action. Two features make them unlike everything else in Prelims. First, they conventionally carry no negative marking, so leaving one blank is pure loss. Second, more than one option may be defensible, and partial credit is awarded — so your job is not to find the single "correct" answer but the most defensible one. The graders are looking for the same qualities GS-IV tests: legality, proportionality, responsibility and a willingness to act rather than evade.

How to rank the four options — a reliable ordering
RankType of optionWhy it scores that way
BestLawful, addresses the substance, keeps a record, and involves the competent authorityShows responsibility with due process — exactly what the paper is screening for
GoodLawful and well-intentioned but incomplete — solves the immediate issue without preventing recurrenceEarns partial credit; nothing wrong with it, just less complete
WeakPassive — refer it upward with no view, or wait and seeEvasion reads as inability to decide; low but rarely zero
WorstIllegal, harsh, public humiliation of a subordinate, or dramatic gestures like resigning or leakingScores zero even when the motive is good, because the means are indefensible
🧠 Four tests to apply in ten seconds
Legality — does any option break a rule? Eliminate it, whatever its benefits. Proportionality — is the response matched to the problem, or is it a sledgehammer? Ownership — does the option have you doing something, or only forwarding the problem? Sustainability — does it stop the situation recurring? An option that passes all four is the answer. And remember the section-specific rule: since there is no negative marking here, mark every one of these questions even when you are torn between two options.

Interpersonal and Communication Skills

  • The communication process and where it breaks. Sender → encoding → message → channel → decoding → receiver → feedback, with noise capable of degrading any stage. The examinable insight is that most administrative communication failures are decoding failures caused by jargon, or feedback failures because no one checked whether the instruction was understood. "I issued the circular" is not communication; confirming that the field office acted on it is.
  • Barriers, by type. Physical (distance, poor connectivity), semantic (jargon, translation, technical vocabulary), psychological (fear of a superior, prejudice, defensiveness), organisational (too many hierarchical layers, information hoarding as power) and cultural or linguistic. Each has a distinct remedy, which is why classifying the barrier is more useful than listing barriers.
  • Active listening is the skill most tested in disguise. Its components: attend without interrupting, paraphrase back to confirm understanding, ask clarifying rather than leading questions, acknowledge the emotion before addressing the content, and summarise the agreed action at the end. In a public hearing this is what converts a shouting match into a record of grievances.
  • Assertive, not aggressive or passive. Passive communication concedes to avoid conflict; aggressive communication wins the exchange and loses the relationship; assertive communication states the position and the reason clearly while respecting the other party. In CSAT options, the assertive choice is almost always the highest-scoring one — "I explained why the request could not be granted and offered a lawful alternative" beats both silence and confrontation.
  • Emotional intelligence and empathy reappear here from GS-IV, and the same caution applies: empathy means understanding the other person’s position, never conceding an unlawful demand. An option that resolves the situation by quietly bending a rule to keep everyone happy is designed to be rejected.

Key takeaways

  • CSAT is qualifying at 33% only: 80 questions, 200 marks, 120 minutes, one-third negative marking. You need roughly 27 clean correct answers. Marks above 66 add nothing to your rank, but falling below 66 makes an excellent GS Paper-I score irrelevant.
  • Three wrong answers cancel one correct, so blind guessing is mathematically negative — but the decision-making section conventionally has no negative marking and awards partial credit, so never leave those blank.
  • In comprehension the answer is always inside the passage. Read the passage first, state the central claim in your own words, then eliminate options by defect: too extreme, too narrow, too broad, outside the passage, or reversed.
  • Qualifier words are load-bearing. A passage saying a policy "may reduce" inequality does not support an option claiming it "eliminates" inequality — that substitution is among the most frequent traps.
  • "Most logical and rational inference" means what follows necessarily and minimally — not what is true, not what you agree with. Prefer the modest inference; a passage describing a problem rarely licenses a specific policy prescription.
  • Rank decision-making options: best = lawful, addresses the substance, records it, involves the competent authority; good = lawful but incomplete; weak = passive referral or wait-and-see; worst = illegal, harsh, humiliating or dramatic (resigning, leaking) even with good motives.
  • Four ten-second tests for decision-making: legality, proportionality, ownership (are you acting or only forwarding?), sustainability (does it prevent recurrence?).
  • Communication: sender → encoding → message → channel → decoding → receiver → feedback, degraded by noise. Most administrative failures are decoding failures (jargon) or feedback failures (nobody confirmed understanding). Issuing a circular is not communication; confirming action on it is.
  • Be assertive, not passive or aggressive: state the position and reason clearly while respecting the other party. In CSAT options the assertive choice almost always scores highest, and empathy never means conceding an unlawful demand.

Practice questions (8)

Attempt each one before opening the answer. Every explanation names the tempting wrong option as well as the right one, because that is where marks are lost.

  1. Approximately how many questions must a candidate answer correctly, with no wrong answers, to clear CSAT?

    1. About 15 of 80
    2. About 27 of 80
    3. About 40 of 80
    4. All 80
    Show answer

    Answer: B — About 27 of 80

    66 marks are needed (33% of 200) and each question carries 2.5 marks, so 66 ÷ 2.5 ≈ 27 clean correct answers. Since three wrong answers cancel one correct at one-third negative marking, a candidate who guesses widely may need considerably more correct answers than 27.
  2. Which section of CSAT should never be left blank, and why?

    1. Data interpretation, because the questions are easy
    2. Decision making, because it conventionally carries no negative marking and awards partial credit
    3. Comprehension, because passages are short
    4. Basic numeracy, because it is Class X level
    Show answer

    Answer: B — Decision making, because it conventionally carries no negative marking and awards partial credit

    With no negative marking and partial credit for defensible options, leaving a decision-making question blank is pure loss. Every other section carries one-third negative marking, so blind guessing there is mathematically negative.
  3. A passage states that a subsidy "may somewhat reduce" rural distress. Which option is most likely to be correct?

    1. The subsidy eliminates rural distress
    2. The subsidy may contribute to easing rural distress, though not decisively
    3. Subsidies are always the best instrument of rural policy
    4. The government should immediately double the subsidy
    Show answer

    Answer: B — The subsidy may contribute to easing rural distress, though not decisively

    Option 1 replaces "may somewhat reduce" with "eliminates" — the classic qualifier-substitution trap. Option 3 is too extreme ("always", "best"). Option 4 imports a policy prescription the passage does not license. Only option 2 preserves the passage’s own hedging.
  4. In a decision-making question, which option would score lowest even if the candidate’s motive is good?

    1. Address the issue, record it in writing and inform the competent authority
    2. Resolve the immediate problem without preventing recurrence
    3. Refer the matter upward without expressing any view
    4. Publicly reprimand the subordinate in front of the office to set an example
    Show answer

    Answer: D — Publicly reprimand the subordinate in front of the office to set an example

    Public humiliation is disproportionate and damages the working relationship, so it scores lowest despite an intention to enforce discipline — the means are indefensible. Option 3 is weak but merely passive, and option 2 earns partial credit for being lawful though incomplete.
  5. A District Officer issues a detailed circular but field offices do not act on it. In communication terms this is most likely a failure of:

    1. Encoding only
    2. Decoding and feedback — jargon-heavy drafting plus no confirmation that the instruction was understood and acted on
    3. The channel alone
    4. No failure — issuing the circular completes communication
    Show answer

    Answer: B — Decoding and feedback — jargon-heavy drafting plus no confirmation that the instruction was understood and acted on

    Communication is only complete with feedback. Most administrative breakdowns are decoding failures caused by jargon combined with the absence of any loop confirming that the instruction was understood and implemented. Option 4 is precisely the misconception the question tests.
  6. Which response is assertive rather than passive or aggressive?

    1. Granting the irregular request to avoid an argument
    2. Explaining clearly why the request cannot be granted and offering a lawful alternative
    3. Refusing curtly and ending the conversation
    4. Saying nothing and letting the file lapse
    Show answer

    Answer: B — Explaining clearly why the request cannot be granted and offering a lawful alternative

    Assertiveness states the position and the reason while respecting the other party. Option 1 is passive, option 3 aggressive and option 4 evasive — and in CSAT options the assertive choice almost always scores highest.
  7. Consider:
    (1) Scoring well above 66 in CSAT improves your final rank.
    (2) Failing CSAT makes an excellent GS Paper-I score irrelevant.
    Which is correct?

    1. Only 1
    2. Only 2
    3. Both 1 and 2
    4. Neither 1 nor 2
    Show answer

    Answer: B — Only 2

    CSAT is purely qualifying, so marks above the 33% threshold add nothing to the rank — statement 1 is false. But falling below it disqualifies the candidate regardless of GS Paper-I performance, which is why the paper deserves enough practice to be safe and no more.
  8. For a short CSAT comprehension passage, the recommended approach is to:

    1. Read the questions first and then scan the passage for keywords
    2. Read the passage first, state the central claim in your own words, then evaluate the options
    3. Answer from prior knowledge of the subject
    4. Always pick the longest option
    Show answer

    Answer: B — Read the passage first, state the central claim in your own words, then evaluate the options

    Keyword scanning works for long passages with detail-based questions; CSAT passages are short and the questions are usually about the argument as a whole, so scanning makes you miss it. Option 3 is the single most common cause of error — the instruction "on the basis of the passage alone" is literal.