Statement-Based Reasoning

A short situation is given and you are asked what you should do, or which response is the most emotionally intelligent. Four reactions follow, and one of them is clearly the most professional. In the 2026 papers this is the biggest family of the chapter, and it comes in four shapes: a workplace or ethics case (an error, a conflict, a found wallet), an emotional-intelligence case (a sad friend, an angry citizen), a problem-solving case (a problem is described and you pick the method that fixes it) and a competing-duties case (an exam is near and a friend needs help).

The logic behind all four shapes is the same. The examiner checks whether your reaction is honest, timely, constructive, within your role and proportionate. Five wrong families keep coming back: hiding, blaming, ignoring, waiting passively and over-reacting.

[[fig:0]]

Picture a staircase from worst to best. Hide, ignore or blame sit at the bottom; then wait and see; then a partial fix; and at the top, inform, correct and prevent. Cross out the lowest rungs first, and from what remains pick the more complete and less harmful reaction. When two options both look good, take the one that is more complete and less harmful.

Two more ideas help. First, size the response to the problem. A small or unintended fault is met with a private, polite word, while fraud, a crime or danger to many people goes to the right authority at once. Second, in emotional cases the first duty is to the person: acknowledge the feeling, then help. A legal-principle item (a doctor discloses patient data by mistake) is answered by the principle whose duty was breached, here confidentiality. All of this is an exam convention and not a law, so read all four options before choosing.

You are given a statement and two (sometimes three) conclusions, and the options are Only I, Only II, Both or Neither. The same job is also worded as "judgements" or "inferences", and some papers give one statement with four sentences to choose from. In every form the question is the same: is this claim forced by the statement?

A conclusion may be in the statement's own words, in different words with the same meaning (all metro stations are under CCTV, so activity at metro stations is monitored), or one plain step away (25 students and 15 boys, so 10 girls). Treat the statement as true even if it clashes with real life.

[[fig:0]]

Think of the statement as a closed box. A valid conclusion is already inside the box or can be reached from inside by one obvious step. If you must step outside and borrow something from common sense, the conclusion is out. A conclusion is never stronger than the statement and never brings anything new.

The method has four short steps. First, underline the strength words in the statement. Second, underline the strength word in the conclusion; if it is stronger, reject it. Third, look for any noun, number or reason in the conclusion that the statement does not contain; if there is one, reject it. Fourth, what is left is the same meaning or one step of simple arithmetic: accept it.

The latest papers add an exam habit: they reward the mild, purpose-matching conclusion. If the government gives free education to girls, "this step will help raise girls' literacy" follows, because it only says what the step is meant to do. The conclusion that fails is the extreme one ("parents will stop sending boys to school", "all industries will shut down"), the one about a different group (boys, rural areas), or a guessed motive.

Sub-case: "Either I or II". Sometimes I and II are exact opposites about the same thing ("the loan will be sanctioned" and "will not be sanctioned") and the statement leaves it open. Then one of them must be true, because there is no third case.

[[fig:1]]

"Either" needs the same subject, the same time, exact opposites and no third case, and the statement must not decide the matter. If it already decides, only that conclusion follows. If a third case exists (a good batsman and a good bowler can both be true or both be false; a price may rise, fall or stay), the answer is "Neither".

Two shapes appear. In the first, a statement is followed by inferences I and II (options Only I, Only II, Both, Neither). In the second, a statement is followed by four sentences and you pick the one that can be inferred. Older sets use a three-way verdict: an inference is definitely true (forced by the text), definitely false (the text contradicts it) or cannot be said (the text is silent). Some sets add "probably true" for likely but not forced.

[[fig:0]]

Inference runs on the same engine as conclusion: put the statement in a box and ask what the box holds. Tick only what the box holds, never what common sense adds. An inference earns a tick if the box contains it word for word, in the same meaning, or by arithmetic on the given numbers. If the box contradicts it, it is false. If the box is silent, it "cannot be said".

Numbers allow calculation but nothing extra. If a library has 4,000 books and 1,500 are in Hindi, Hindi books are fewer than half, but you cannot say how often they are borrowed. No percentage can be invented from data that was not given.

Inference gives you two small freedoms that conclusion does not stress. A statement can be turned around softly: "reducing screen time before bed improves sleep" allows "excessive screen time before sleep may harm sleep". And a meaning built into a word may be used: luxury items are non-essential goods. What is never allowed is a new fact about a person (parents are actors, has done Hollywood films) or a sweeping word ("all sleep problems are caused by screens").

Three checks complete the method. The converse trap: "all mammals are warm-blooded" does not give "all warm-blooded animals are mammals". One event is not a habit: overtime last week is not "always works overtime". And "every Monday" is not "only on Monday".

An advertisement, an instruction, a piece of advice, an announcement or a government decision is followed by two or three assumptions, and you decide which are implicit. An assumption is a belief the speaker must already hold for the statement or action to make sense. It is a hidden pillar under a bridge: you do not see it, but without it the bridge falls. It need not be true in real life, it is never the statement repeated, and it is never a result of the statement.

[[fig:0]]

Ask what the speaker believes about the world, the audience and the plan. A scheme assumes funds and public cooperation, an advertisement assumes that people will read and respond, and advice assumes that the listener can follow it.

The method is the Negation test. Step 1: find the action and its purpose. Step 2: flip the option ("schools have vacancies" becomes "schools have no vacancies"). Step 3: if the action now looks pointless or impossible, the option is an assumption; if nothing changes, it is not. If denying it makes the statement collapse, it is an assumption.

The test also settles the tricky cases of the latest papers. "Permanent work from home to improve productivity": deny "employees were less productive earlier" and the company can still want better productivity, so that option is not implicit. Deny "work from home will give better productivity in the long run" and the decision is pointless, so that one is implicit. Likewise "students should do two hours of physical activity": deny "students now do too little" and the rule loses its reason, so it is implicit.

Use quick rejects first: a contradiction, a new actor or motive, a strict word the statement does not need (all, only, never, must), a mere repeat of the statement, or a result. Apply the test only to the survivors. Each of I, II (and III) is tested separately.

The statement describes a problem, and the courses of action are the steps an authority might take. The question comes in two shapes: two or three numbered courses (1, 2 or I, II) with options such as Only 1, Only 2, Both, Neither; or a "Judgement: What should be done / what is the most effective action?" with four steps of which you pick one.

A good course of action is real and immediate, within the power of the actor named, proportionate, and it actually fixes or contains the problem. Simple problems (poor reading habits, malnutrition, unemployment) need steady improvement steps. Complex or urgent ones (a riot, an epidemic, a collapse, a flood) need immediate, concrete action.

[[fig:0]]

The method is the four gates. P is the problem and each course must pass four tests in order. Gate 1: is it real and immediate? ("After six months" or "within a week" for a huge job fails.) Gate 2: is it within the power of the actor? (A district officer cannot redesign the national railway.) Gate 3: is it proportionate and targeted, with no harm to innocents? Gate 4: does it fix or contain the problem rather than shift or worsen it? Keep only the options that pass all four gates.

[[fig:1]]

For punishments, use the harshness ladder: 1 warning, 2 fine, 3 repair or restriction, 4 suspension, 5 dismissal or arrest. Take the lowest rung that solves the problem; the top rung is justified only after repeated violation.

Valid shapes: send an expert team or fact-finding committee, treat the cause, inform the public, compensate victims, detain a clearly guilty person, inspect and fine rule-breakers, and run awareness drives alongside. Invalid shapes: blame, resign, ban everything, close the facility, destroy a whole stock because part of it is bad, punish the victims, or a step that attracts more of the problem. "Enforce the existing law" is valid when it names a specific step.

A post, its conditions (age, education, marks, experience, licence) and usually a cut-off date are given, followed by one case or four candidates. The options are either names, or "Eligible / Not eligible (reason) / Information insufficient". The newest form is the criteria with an exception: "if condition (1) is not met, but the candidate has …, refer to the CEO", and the options become Select, Reject or Refer to an authority. The skill here is care, not cleverness.

[[fig:0]]

The method is the Tick-Cross grid. Write the conditions c1, c2, c3 across the top and the candidates down the side. Go candidate by candidate. The first violated condition gets a cross and you stop there. The only candidate with no cross is the answer. One broken condition ends the candidate.

Date work needs care. Count the age on the given cut-off date, not today. A person who turns a new age after the cut-off date is still the younger age on it. "21 to 30 years (both inclusive)" allows both 21 and 30. "More than 2 years" excludes exactly 2, while "at least 2 years" and the sign ≥ include it.

[New Pattern] The exception clause. An exception is a side door, and it opens only when three things hold together: (1) the condition that failed is exactly the one the exception names; (2) every other condition is still met; (3) the exception's own requirement is met in full (15 years of experience means 15 or more, and the named degree must be held). Then the answer is "Refer to" that very authority, so read its name in the options. If any one of the three fails, the answer is Reject. If nothing failed, it is Select and the exception never comes into play.

[[fig:1]]

Two cautions. Do not reject a candidate on a condition the post never listed. And if a needed fact is missing (night-shift willingness is not stated), the case cannot be decided, and a missing fact never means "yes".

Three shapes live in this type. Link items give Statement I and Statement II (or A and B) with the options I causes II, II causes I, independent causes, or a common cause. Reason items describe a situation, often a surprising one, and ask for the most probable reason (or which of two reasons P and Q can explain it). A rare third shape is the Assertion and Reason pair with four standard codes. All three rest on one idea: an explanation needs a believable mechanism and must never break a fact given in the question.

Link items. A cause comes before its effect, and there must be a believable mechanism: you can say "because of I, II happened". Sequence alone is not cause ("the bell rang and then it rained"). First look for a mechanism, then for a hidden common cause, and only then call the events independent. One occasion behind two same-direction events is a common cause (a holiday closes both schools and offices); separate reasons give independent effects.

[[fig:0]]

Remember the remedy check. A response is the effect of the trouble it answers: a road-safety campaign follows a rise in accidents, so the trouble is the cause. But if the second statement reports the result after the step (delivery times improved after a new bonus scheme), the step is the cause.

Reason items. List the facts and ask of each option: "If this were true, would the whole picture happen, with every fact still true?" Words like "despite", "although" and "even though" mark the facts that must stay true. Crops dried up despite irrigation systems: the systems exist, so the failure lies in delivery, and poor maintenance of the channels fits. Choose the most probable of the given options, the one with the more direct mechanism.

[[fig:1]]

Assertion and Reason. This is a fact-check: judge A alone, then R alone, then say "A, because R".

[[fig:2]]

A false A kills codes (a), (b) and (c); a false R kills (a), (b) and (d). If both are true and R is the real cause, the code is (a); if R is a true but unrelated fact, it is (b). Example: "Sodium is stored under kerosene" (A) and "sodium conducts electricity" (R) are both true, but R is not the reason, so the code is (b). A short fact bank is in the Kavach Points.

Two shapes appear. Argument items give a "Should …?" statement and Arguments I and II (usually one Yes and one No), and you pick which is strong. Strengthen and weaken items [New Pattern] give a plan or claim and ask "which, if true, best supports / most weakens it?", with four extra facts as options. Both shapes judge the quality of the reasoning, not whether you personally agree.

A strong argument is relevant to the exact question, factual or verifiable, and shows a real consequence for the people concerned (health, safety, rights, the economy, fairness). A weak one is a bare claim, an emotion, a personal trait, a comparison with another country, popularity, or an absolute such as "only". The method is the 3-Gate filter: right question, real consequence, free of "only", comparison and emotion. Pass all three gates and the argument is strong.

[[fig:0]]

Judge each argument on its own. In the latest CHSL papers a Yes and a No were both strong, because each named a real concern (harm to society from misinformation, and the right to free speech).

For strengthen and weaken items, shorten the plan to "action → aim" (electric buses → less pollution) and test the link. [[fig:1]] An option strengthens if it shows that the problem is real and the action hits its main source ("most coastal waste is single-use plastic" supports a ban on single-use plastic), or that the action can actually work. It weakens if it shows the cause lies elsewhere ("most dust comes from construction sites, not the main road"), that a side effect defeats the aim, or that something more urgent competes (poverty and healthcare against space programmes). Ignore taste, other plans and unrelated facts, and read the stem word first, since strengthen and weaken are opposites.