Data sufficiency is not about solving a problem — it is about deciding whether a problem can be solved given certain pieces of information. That is a subtle but critical distinction, and the candidates who miss it in the exam hall lose marks on questions they would otherwise get right.
Here is the core idea: you are given a question and two statements. You do not answer the question itself. You answer a meta-question — does the information in Statement I alone, Statement II alone, both together, or neither allow you to reach a unique, definitive answer?
Think of it like this. A doctor asks, "Does this patient have diabetes?" Statement I says, "The patient's fasting glucose is above 126 mg/dL." That alone gives a definitive yes — you do not need Statement II at all. But if Statement I said only "the patient is overweight," you cannot conclude anything definitive. You need more data.
That is the logic of data sufficiency: you are the analytical officer deciding whether the case file has enough information before escalating further. You do not solve the crime — you decide if the evidence is sufficient.
The reason UPSC tests this is well-grounded. Civil servants routinely encounter situations where they must evaluate whether available information is adequate for a decision. Data sufficiency directly tests that capacity.
A few foundational rules before you proceed:
A unique answer is always required. If Statement I produces two or more possible answers to the question, it is insufficient — even if both answers are "reasonable." Sufficiency means one and only one answer follows.
"No" is a valid definitive answer. If Statement I lets you conclude with certainty that the answer to the question is "No," that statement is sufficient. Many candidates wrongly think sufficiency means the answer must be "Yes."
Do not carry information from one statement into the other when testing individually. When you test Statement I alone, Statement II does not exist. When you test Statement II alone, Statement I does not exist. Only when testing both together do you combine them.
UPSC presents data sufficiency questions in one of two formats, but the underlying logic maps to these four standard outcomes:
Your job on every question is to determine which of these four situations applies.
Pass 1: Test Statement I alone. Ignore Statement II entirely. Ask: given only Statement I (plus the information in the question stem), can the question be answered with exactly one answer? If yes, mark it "I-sufficient." If no, mark it "I-insufficient."
Pass 2: Test Statement II alone. Ignore Statement I entirely. Repeat the same test for Statement II.
Pass 3: Combine, if needed. If both statements were individually insufficient, combine them. Ask: does the combination produce a unique answer? If yes, option (c). If no, option (d).
If Statement I alone was sufficient, check Statement II alone. If Statement II is also sufficient on its own, the answer is (b). If only one was sufficient, the answer is (a) with a note of which one.
Arithmetic/Algebraic questions: You are given partial information about prices, ages, scores, or quantities and must decide if the given statements pin down a unique numerical value.
Look for: the number of unknowns versus the number of independent equations or constraints the statements provide. One unknown typically needs one independent equation. Two unknowns need two independent equations. This is not the only check, but it is a fast screening tool.
Yes/No questions: The question asks something like "Does P score more than Q?" or "Is X divisible by 3?" Here sufficiency means the statement allows you to answer definitively — either always yes or always no — for every possible value consistent with the given conditions.
This is the type where most candidates go wrong. They test one specific numerical example, get a "Yes," and assume the statement is sufficient. The correct test: is there any scenario consistent with the statement that gives the opposite answer? If yes, the statement is insufficient.
Combinatorial or arrangement questions: Here you are checking whether the constraints uniquely fix a configuration (seating, ordering, grouping). These are among the hardest because the "constraint space" is less obvious.
When both statements together appear to contradict each other, resist the urge to conclude (d) too quickly. Re-read the question stem — sometimes the apparent contradiction means the question's premise is impossible, which actually means both statements together do resolve the question (the answer is "the situation cannot arise"). UPSC rarely sets outright logical contradictions, but be alert.
When the question has n unknowns, count the number of genuinely independent constraints provided by the statement. If a statement provides n independent constraints for n unknowns, it is sufficient (assuming no degenerate cases like parallel lines in a two-variable system). If it provides fewer, it is insufficient. This reduces many DS questions to a 10-second constraint count rather than a full algebraic solution.
Consider a question asking for the price of article p. Suppose Statement I says p ≤ r and the problem constrains p + r = 34 with both values above 16. Statement I alone allows p = 17, r = 17 or p = 16, r = 18 and so on — multiple solutions exist. Statement I alone is insufficient. Statement II (r ≤ p) is similarly insufficient. But together, p ≤ r and r ≤ p means p = r, and p + r = 34 gives p = r = 17. One unique answer — sufficient together.
This is the pattern tested directly in PYQ 6a205bf5ccfcbabc57a3a5d9. The key cognitive move: look for what a pair of inequalities does when stacked — they often force equality.
Use a strict three-pass sequence before marking any answer. Pass 1: test I alone (30 seconds max). Pass 2: test II alone (30 seconds max). Pass 3: combine only if both fail. This prevents the most common error — combining statements subconsciously during Pass 1.
Micro-example: Question asks "Is n even?" Statement I: n² is even. Test I alone — if n² is even, n must be even (since odd² is odd). Statement I alone is sufficient. You never need Pass 2. Standard approach of testing both and then combining: ~90 seconds. Three-Pass Funnel with early exit: ~20 seconds.
For yes/no questions, try to construct one scenario consistent with the statement where the answer is "Yes" and one where the answer is "No." If you find both, the statement is insufficient — immediately. You do not need to map the whole solution space.
Micro-example: Statement I says "m is not a prime number." Question: "Did X receive a coin of denomination 5?" Build a case without 5 that gives composite m (e.g., 1+2+10 = 13 — prime, 1+2+20 = 23 — prime, 1+10+20 = 31 — prime, 2+10+20 = 32 — composite). Build a case with 5 that gives composite m (e.g., 1+2+5 = 8 — composite). You now have two composite totals, one with and one without a 5-coin. Statement I is insufficient. Time to build two counterexamples: ~40 seconds. Full enumeration of all cases: ~3 minutes.
When two statements each give an inequality about the same pair of variables (e.g., p ≤ r and r ≤ p), they always force equality when combined. This is a one-second recognition pattern — you do not need to substitute values.
If Statement I gives X ≤ Y and Statement II gives Y ≤ X, the combined answer is always X = Y. Immediately check if this equality plus the question stem gives a unique numerical answer. If it does, mark (c) without further work. Standard algebraic substitution and testing: ~90 seconds. Inequality Stack recognition: ~5 seconds.
Before computing anything, count the unknowns n in the question and count the independent constraints provided by each statement. If a single statement provides n independent constraints for n unknowns (with no redundancy), flag it as "likely sufficient" and verify in 10 seconds. If it provides fewer, it is insufficient — no calculation needed.
Micro-example: "What is price of p?" gives one unknown. Statement I: p ≤ r (an inequality, not an equation — 0 independent equations). Immediately classified insufficient without any substitution. Saves 45 seconds per question where this screen triggers.
Option (d) — "cannot be answered even with both statements" — is the answer when you can always build two valid scenarios from the combined statements that give opposite answers. Do not mark (d) just because the question feels hard. The correct trigger is: after combining both statements, demonstrate two specific numerical examples — one where answer is "Yes" and one where answer is "No" — both consistent with all given conditions.
Micro-example from PYQ 6a205bf5ccfcbabc57a3a5db: Even with both statements, setting P=30, Q=20, R=20, S=30 satisfies both (P+Q = R+S and P+S > Q+R) and gives P > Q. But setting P=25, Q=25, R=15, S=35 satisfies both statements and gives P = Q. Two valid scenarios, two different answers — option (d). Building these two examples: ~60 seconds. Much faster than attempting full algebraic proof of insufficiency.
In the exam hall, follow this decision path without deviation:
Step 1 — Read the question stem carefully. Identify what type of answer is needed: a specific number, a yes/no, or a ranking. Extract all information embedded in the stem — constraints given there apply to both statements always.
Step 2 — Test Statement I alone (cap: 40 seconds). Attempt to reach a unique answer using only Statement I. If you reach a unique answer, mark I as "sufficient" and move to Step 3. If you find two valid answers, mark I as "insufficient" and move to Step 3.
Step 3 — Test Statement II alone (cap: 40 seconds). Repeat the test for Statement II in isolation.
Step 4 — Combine if needed. If both were insufficient, combine them (cap: 60 seconds). Reach unique answer = option (c). Cannot reach unique answer = option (d).
Step 5 — Map to option. I sufficient, II insufficient = (a) referencing I. I insufficient, II sufficient = (a) referencing II. Both sufficient alone = (b). Neither alone but together sufficient = (c). Neither, even together = (d).
Total cap per question: 2.5 minutes. If you exceed this, make your best call and move on — UPSC CSAT time pressure is real.
Why this question (6a205575ccfcbabc57a3a305): This tests whether candidates can enumerate a small combinatorial space systematically rather than guessing. It also tests the critical skill of verifying both possible answers (with and without a 5-coin) before marking a statement sufficient.
Solving path: First, list all combinations of three coins from {1, 2, 5, 10, 20} — there are C(5,3) = 10 combinations. Separate them into two groups: those containing a 5-coin (6 combinations) and those not containing a 5-coin (4 combinations). The four without a 5: {1,2,10}=13, {1,2,20}=23, {1,10,20}=31, {2,10,20}=32. The six with a 5: sums are 8, 16, 26, 17, 27, 35. Now test Statement I (m is not prime): composites without 5 → only 32. Composites with 5 → 8, 16, 26, 27, 35. Multiple composites exist in both groups, so you cannot determine from Statement I alone whether a 5-coin was included. Statement I is insufficient. Now test Statement II (digit sum of m > 5): digit sums without 5 → 1+3=4, 2+3=5, 3+1=4, 3+2=5. None exceed 5. Digit sums with 5 → 8, 1+6=7, 2+6=8, 1+7=8, 2+7=9, 3+5=8. All exceed 5. So if digit sum > 5, X must have received a 5-coin. Statement II alone is sufficient. Answer: (a), Statement II alone.
Why this question (6a2059edccfcbabc57a3a537): This question tests the ability to recognize what information is still missing even when statements seem detailed. Candidates often combine statements mentally and arrive at (c) too quickly — the official answer is (c) but the reasoning path requires careful handling of what "behind by 3 goals" means at the start of the last 10 minutes.
Solving path: Let Q's score before the last 10 minutes = x, so P's score before last 10 minutes = x − 3. Statement I alone: P scored 4 in last 10 minutes. P's final = x − 3 + 4 = x + 1. But Q's final score is unknown — Q could have scored anything in the last 10 minutes. P's final could be more or less than Q's final. Insufficient alone. Statement II alone: Q's total = 4. Q started with x, ended at 4. But we do not know P's scoring in the last 10 minutes. P's final could be more or less than 4. Insufficient alone. Both together: P's final = x + 1, Q's final = 4. We need x. The intended reading by UPSC assumes the "behind by 3" is at the start of the last 10 minutes with Q having scored exactly 4 up to that point (x = 4), meaning P had 1, then P scored 4 more to reach 5 vs Q's total of 4. P wins. This gives the official answer (c) — both statements together are sufficient under the standard intended interpretation.
Why this question (6a205bf5ccfcbabc57a3a5db): This is the most analytically demanding type in UPSC DS — it requires proving that even the combination of both statements cannot yield a unique answer. Candidates who only try one set of numbers and get "P > Q" will incorrectly mark (c). The skill here is constructing two valid counterexamples.
Solving path: Let P, Q, R, S be the marks. Statement I: P + Q = R + S. Statement II: P + S > Q + R. Test Statement I alone: set P=30, Q=20, R=20, S=30 (P+Q=R+S=50). Here P > Q. Set P=20, Q=30, R=30, S=20 (P+Q=R+S=50). Here P < Q. Two valid scenarios, opposite answers. Insufficient. Test Statement II alone: set P=30, Q=20, R=10, S=5 (P+S=35 > Q+R=30). Here P > Q. Set P=25, Q=20, R=1, S=100 (P+S=125 > Q+R=21). Here P > Q — can we build a "P < Q" case? Set P=10, Q=20, R=1, S=100. P+S=110 > Q+R=21. Here P < Q. Insufficient. Combine both: Need P+Q = R+S and P+S > Q+R. From Statement II: (P+S) − (Q+R) > 0 → (P−Q) > (R−S). From Statement I: P+Q = R+S → P−R = S−Q → P−Q = R−S. If P−Q = R−S, substituting into the inequality: P−Q > R−S becomes P−Q > P−Q, a contradiction. This seems to make no scenario possible — but the problem states these are given as facts, which means our algebra shows the statements cannot simultaneously hold unless P−Q > P−Q, which is impossible. However, UPSC's official answer is (d), meaning the statements are treated as individually inadequate and together still insufficient, because the question is asking for a definitive yes/no — and the algebraic contradiction means no scenario simultaneously satisfies both, making any answer derived from combining them vacuously undefined. The key takeaway: when statements appear to contradict each other algebraically, mark (d).
Why this question (6a205bf5ccfcbabc57a3a5d9): Clean illustration of the Inequality Stack pattern — two opposing inequalities force equality, and equality plus a known sum gives a unique answer.
Solving path: From the stem: p + q + r = 50, q = 16 (least price), so p + r = 34. Since q is the least, both p ≥ 16 and r ≥ 16. Statement I alone (p ≤ r): p could be 16 with r = 18, or p = 17 with r = 17. Multiple values for p. Insufficient. Statement II alone (r ≤ p): r could be 17 with p = 17, or r = 16 with p = 18. Multiple values for p. Insufficient. Both together: p ≤ r (from I) and r ≤ p (from II) forces p = r. With p = r and p + r = 34, we get p = r = 17. Unique answer. Answer: option (c).
Carrying information between statements. When testing Statement I alone, many candidates subconsciously recall something from Statement II and use it. Maintain strict isolation — cover Statement II physically if you have to.
Treating "I can find an answer" as "the statement is sufficient." If you substitute a specific value and get an answer, that is one data point. You have not proven sufficiency. You must confirm no other value consistent with the statement gives a different answer.
Marking (d) when the question feels "unsolvable" without actually testing. Option (d) requires demonstrating that even the combination fails. Do not assume difficulty equals insufficiency.
Ignoring information in the question stem. Constraints given in the stem — like "q is the least price" or "X received exactly three coins" — are always in play and often provide the extra constraint that makes a statement sufficient. Missing these is a frequent source of error.
Confusing a definitive "No" with insufficiency. If Statement I allows you to conclude with certainty that the answer is "No," that statement is sufficient. "Sufficient" does not mean the answer is "Yes."
Algebraic contradiction between statements not being flagged as option (d). When combining both statements produces a logical impossibility (as in PYQ 6a205bf5ccfcbabc57a3a5db), some candidates pick option (c) thinking the impossibility resolves the question. It does not — option (d) is correct in such cases.