By the end of this chapter you'll be able to…

  • 1State the common shape of every analytical question
  • 2Choose a representation before beginning to reason
  • 3Distinguish deduction from induction and what each question asks for
  • 4Explain why validity and truth are independent
  • 5State the valid conversion of each of the four syllogism forms
  • 6Recognise premise combinations that yield no conclusion
  • 7Find the counter-arrangement that breaks a candidate conclusion
  • 8Recognise when an either-or conclusion pair is valid
  • 9Apply the negation test to identify an assumption
  • 10Distinguish an assumption from a conclusion and a course of action
  • 11Locate the gap between evidence and conclusion in an argument
  • 12Name the standard reasoning gaps and their typical weakeners
  • 13Identify affirming the consequent and denying the antecedent
  • 14Draw the frame and order constraints by restrictiveness in an arrangement puzzle
  • 15Apply the facing convention correctly in circular arrangements
  • 16Distinguish 'between' from 'immediately between'
  • 17Resolve blood relations outward from the speaker one step at a time
  • 18Track relative turns correctly in a direction problem
  • 19Compute net displacement using the standard Pythagorean triples
  • 20Compute the angle between clock hands from hours and minutes
  • 21Reduce calendar problems to odd days and apply the leap-year exception
  • 22Evaluate data-sufficiency statements in complete isolation
💡
Why this chapter matters in GATE
Every question in this area has the same underlying shape: you are given a set of constraints and the answer is whatever survives all of them. Nothing is discovered by insight; things are eliminated by rules. That reframing has a practical consequence, because the difficulty of an analytical question is almost entirely a question of representation. A seating puzzle written out in sentences is hard, and the same puzzle drawn as a circle with eight positions is easy. A blood-relation chain described in words is hard, and drawn as a family tree it is trivial. So the discipline is to choose the diagram before doing any reasoning, place the most restrictive constraint first, and let the rest eliminate. The second principle is that validity is not plausibility: a conclusion follows only if it must be true whenever the premises are, and whether it is sensible in the real world is exactly what the wrong options appeal to.

Before you start — revise these

🔗
Verbal Aptitude
Verbal deduction and the discipline of rejecting plausible but unsupported options are developed there and apply unchanged to analytical questions.
🔗
Discrete Mathematics
Propositional logic, implication and the standard fallacies are formalised there; this chapter uses them in their verbal, exam-facing form.

Analytical Aptitude

Analytical Aptitude is the reasoning half of General Aptitude: logic, deduction and induction, analogy, numerical relations and constraint puzzles. It carries part of the 15 marks that every GATE paper allots to General Aptitude, and it is the part most amenable to method.

Every question in this area has the same underlying shape. You are given a set of constraints, and the answer is whatever survives all of them. Nothing is discovered by insight; things are eliminated by rules.

That reframing has a practical consequence. The difficulty of an analytical question is almost entirely a question of representation. A seating puzzle written out in sentences is hard; the same puzzle drawn as a circle with eight positions is easy. A blood-relation chain described in words is hard; drawn as a family tree with two symbols for gender it is trivial.

So the discipline is: choose the diagram before doing any reasoning, place the most restrictive constraint first, and let the remaining constraints eliminate.

The second principle is that validity is not plausibility. In a deduction question, a conclusion follows only if it must be true whenever the premises are true. Whether it is true in the world, or sensible, or what a reasonable person would do, is irrelevant and is exactly what the wrong options appeal to.

1. Deduction and Induction

GATE names both explicitly in its syllabus, and the distinction decides what counts as a correct answer.

Deduction moves from general to particular and preserves certainty. If all A are B and x is an A, then x is a B, with no room for doubt. A valid deduction cannot have true premises and a false conclusion.

Induction moves from particular to general and preserves only likelihood. Observing that every swan examined so far is white supports but does not establish that all swans are white.

The examinable consequence is that a deduction question asks what must follow, while an induction or inference question asks what is best supported. An option that is merely reasonable loses in the first case and can win in the second.

Validity and truth are independent. The argument "All birds can fly; a penguin is a bird; therefore a penguin can fly" is perfectly valid and has a false conclusion, because a premise is false. Questions exploit this by supplying premises that are obviously false in reality, and candidates reject valid conclusions for being untrue.

2. Syllogisms

Four statement forms cover the entire topic, and each permits exactly one conversion.

StatementValid conversion
All A are BSome B are A
No A is BNo B is A
Some A are BSome B are A
Some A are not BNothing

"All A are B" never gives "All B are A". This single confusion accounts for more errors here than everything else combined.

Two structural checks eliminate most questions before any diagram is drawn. Two negative premises yield no conclusion, because negatives only exclude and never establish that anything exists in a shared region. Two particular premises — both beginning with "some" — likewise yield nothing, since two separate overlaps need not touch each other.

When the premises do permit something, the reliable method is to hunt for the arrangement that breaks a candidate conclusion rather than one that supports it. Draw the premises as regions, then push the circles as far apart as the premises allow. A conclusion follows only if it survives every permitted arrangement, so a single counter-arrangement settles it.

Either-or conclusions are the one place two individually invalid conclusions combine into a valid pair. The pair is valid when the two options exhaust the possibilities and neither alone is certain — typically when the premises leave exactly two arrangements, one making the first conclusion true and the other the second.

3. Assumptions, Conclusions and Courses of Action

These three question types look similar and are decided by different tests.

An assumption is something the argument takes for granted and needs in order to work. The test is negation: if denying the candidate assumption destroys the argument, it is an assumption; if the argument survives, it is not.

A conclusion must be derivable from the statement alone. Additional information, however plausible, disqualifies it. This is the same discipline as comprehension: the passage is the universe.

A course of action is judged on whether it addresses the stated problem and is practically implementable. Options that are drastic, that address a different problem, or that assume authority nobody has, fail even when they would work.

A worked distinction makes the difference concrete. Given "The college has decided to install CCTV cameras in all laboratories":

  • An assumption is that the cameras will deter or record the behaviour the college is worried about.
  • A conclusion would be that the college has some concern about laboratory conduct.
  • A course of action would be to also brief students on the policy.

4. Strengthen, Weaken and Flawed Reasoning

Critical-reasoning questions present a short argument and ask what supports or undermines it.

Locate the conclusion first, then the evidence, then the gap between them. Every strengthener closes that gap and every weakener widens it, so naming the gap explicitly makes the options sort themselves.

The commonest gaps repeat and are worth recognising by name.

GapTypical weakener
Correlation treated as causationAn alternative cause
Unrepresentative sampleThe sample differs from the population
Part-to-whole leapWhat holds for a part fails for the whole
Missing comparisonThe control group did as well
Percentage versus absoluteThe base changed

An option that merely restates the conclusion does not strengthen it, and an option that attacks a premise's phrasing rather than the inference does not weaken it. Both are standard distractors.

For flawed-reasoning questions, the answer names the structural error rather than disputing the content. Circular reasoning, denying the antecedent, and affirming the consequent are the three that appear.

Affirming the consequent deserves a note because programmers meet it constantly. From "if the program compiles then the syntax is correct" and "the syntax is correct", nothing follows about compilation.

5. Arrangements and Ordering

Seating, ranking and scheduling puzzles are the most representation-sensitive questions in the section.

Draw the frame first. A row of six seats, a circle of eight, a table with two columns. Then place the most restrictive constraint, which is usually an absolute position or an adjacency, before any relative one.

Three conventions prevent the commonest errors.

In a circular arrangement facing the centre, left and right are reversed relative to how they look on the page. Facing outward, they read normally. Half of all circular-arrangement errors come from this alone, and the fix is to write the direction convention on the diagram before starting.

"Between" does not mean "immediately between" unless the question says so. Treat it as an ordering constraint, not an adjacency, unless the word immediately or exactly appears.

A constraint of the form "A is not adjacent to B" is often more powerful than a positive one, because it eliminates several arrangements at once. Applying negative constraints early rather than last saves substantial time.

For ordering by rank, height or marks, draw a single line and place items relative to each other. Note carefully whether the question counts from the top or the bottom, since "third from the top in a group of nine" is "seventh from the bottom", and the off-by-one is a standard trap.

6. Blood Relations

Draw a family tree with a fixed convention: males as squares, females as circles, a horizontal line for marriage, a vertical line for descent, and siblings on the same horizontal level.

Work strictly from the speaker outwards, one relation at a time. "The father of my son's mother" is resolved by finding my son, then his mother, then her father — three steps, no shortcuts.

Two traps recur. Gender is often unstated and unknowable, and a question whose answer requires knowing whether a person is male or female may be unanswerable by design. And "brother-in-law" and similar terms have several distinct meanings — a spouse's brother, a sister's husband, or a spouse's sister's husband — so a question using them usually admits more than one tree unless it says which.

Coded relation puzzles use symbols such as "A + B means A is the father of B". These are solved by substituting the code definitions into the expression from the innermost operation outwards, exactly as an expression is evaluated.

7. Directions and Distances

Fix north at the top of the page and draw each leg as a vector.

Left and right turns are relative to the direction currently faced, which is why the diagram must be drawn rather than tracked mentally. A person facing south who turns left is facing east, not west.

The final displacement is a straight line from start to finish, so net horizontal and net vertical displacements are computed separately and combined by Pythagoras.

Recognise the standard triples, since almost every such question is constructed around one: 3-4-5, 5-12-13, 8-15-17. If your legs come out as 9 and 12, the answer is 15.

A shadow-based direction question uses one fact: the sun rises in the east and sets in the west, so a morning shadow falls towards the west and an evening shadow towards the east. At noon the shadow is shortest and points north or south depending on hemisphere and season, which is why questions stay away from noon.

8. Coding, Decoding and Numerical Relations

Letter coding works on positional shifts. Knowing the alphabet's positions both forwards and backwards is what makes these fast.

The complement of a letter's position is 27 minus its position, so A pairs with Z, B with Y, and M with N. This "opposite letter" relation appears constantly.

The systematic approach is to write the plaintext and the code one above the other, compute the shift for each letter, and check whether the shift is constant, alternating, or increasing.

For number-and-symbol codes, treat the mapping as a function and test it on every given pair before applying it to the target. A rule that fits two examples and fails a third is not the rule, and testing all given pairs before answering prevents the commonest error in this area.

Numerical-relation questions in GATE frequently take the form of a defined operator: "a * b means a squared minus b". These are solved by direct substitution, working from the innermost bracket outwards, and the only difficulty is order of operations.

9. Clocks and Calendars

Both topics reduce to arithmetic on a fixed rate.

The minute hand gains 5.5 degrees per minute on the hour hand. The hour hand moves 0.5 degrees per minute and the minute hand 6 degrees per minute, so the angle between them at hours and minutes is

taking the smaller of and .

The hands coincide 11 times in 12 hours, not 12, because between 11 o'clock and 1 o'clock they coincide only once, at 12. They are at right angles 22 times in 12 hours.

For calendars, work in odd days, the remainder when a day count is divided by 7. An ordinary year has 365 days and therefore 1 odd day; a leap year has 2.

A century has 5 odd days, and 400 years has 0, which is why the calendar repeats exactly every 400 years.

The leap-year rule is divisible by 4, except centuries, unless divisible by 400. So 1900 was not a leap year and 2000 was, and questions are constructed specifically around this exception.

10. Data Sufficiency

A data-sufficiency question gives a question and two statements, and asks whether each alone, both together, or neither suffices.

The task is to decide whether the answer is determined, not to compute it. Computing wastes time and, worse, tempts you to combine information from both statements while evaluating one.

The discipline is to evaluate statement I completely, then deliberately forget it before evaluating statement II. Carrying information across is the single commonest error, and it produces the wrong answer in exactly the cases the question was designed to test.

A statement is sufficient when it forces a unique answer, not when it produces a plausible one. If two different values are consistent with the statement, it is insufficient, and finding two such values is the fastest way to prove insufficiency.

11. Worked Examples

Example 1. Statements: All engineers are graduates. Some graduates are researchers. Conclusions: (I) Some engineers are researchers. (II) Some researchers are graduates. Which follows?

Take them separately.

Conclusion II is the valid conversion of "Some graduates are researchers", so it follows.

Conclusion I does not. Arrange the regions so that the researcher circle overlaps the graduate circle entirely outside the engineer region. Both premises remain satisfied and conclusion I is false, so it is not forced.

The answer is that only II follows.

The trap is that conclusion I is plausible — in the real world some engineers certainly are researchers — and plausibility is not what the question asks about.

Example 2. Six people A, B, C, D, E, F sit in a row facing north. C is at one end. B is immediately to the right of D. A is third from the left. E is not adjacent to A. Where does F sit?

Draw six positions numbered 1 to 6 from left to right.

A is at position 3. C is at an end, so position 1 or 6.

B is immediately right of D, so DB occupies consecutive positions: (1,2), (4,5) or (5,6).

If DB is at (1,2), then C must be at 6, leaving positions 4 and 5 for E and F. But E cannot be adjacent to A at position 3, so E cannot be at 4, giving E at 5 and F at 4.

Check the alternatives. If DB is at (4,5), then C is at 1 or 6 and E and F take the remaining two of 2 and 6. E cannot be at 2, since 2 is adjacent to 3. So E is at 6, which forces C to 1 and F to 2. This is also consistent.

Two consistent arrangements exist, so the puzzle as stated does not fix F uniquely — which is itself the point worth noticing. In an exam, a second consistent arrangement means a constraint has been misread, and the usual culprit is treating "between" as "immediately between" or forgetting that an end can be either end.

Example 3. Pointing to a photograph, a man says, "She is the daughter of the only son of my grandmother." How is she related to him?

Work outwards from the speaker, one step at a time.

"My grandmother" — the speaker's grandmother.

"The only son of my grandmother" — since she has only one son, and the speaker descends from her, that son is the speaker's father.

"The daughter of my father" — that is the speaker's sister.

She is his sister.

Note what the word "only" is doing: without it, the grandmother's son could be an uncle, and the woman would be a cousin. Removing a single word changes the answer, which is why these questions must be read literally.

Example 4. A man walks 10 km north, turns right and walks 6 km, turns right again and walks 4 km. How far is he from his starting point, and in which direction?

Draw north upward. The first leg is 10 km up.

Turning right while facing north means facing east, so the second leg is 6 km east.

Turning right again means facing south, so the third leg is 4 km south.

Net vertical displacement is 10 north minus 4 south, which is 6 km north. Net horizontal displacement is 6 km east.

The straight-line distance is the hypotenuse of a 6 by 6 triangle, which is , approximately 8.49 km, and the direction is north-east.

The reason the diagram is not optional is the second turn. A person facing east who turns right faces south, and tracking that mentally while also tracking distances is where errors enter.

Example 5. At what time between 3 and 4 o'clock are the hands of a clock at right angles?

At exactly 3 o'clock the hour hand is at 90 degrees and the minute hand at 0.

Use the angle formula with : the angle is , and we need this to equal 90.

Either , giving , which is the trivial 3 o'clock position where the hands are already at right angles.

Or , giving and .

So the hands are again at right angles at about 3:32:44.

Both answers are correct, and a question asking "at what time between 3 and 4" usually intends the second. This is worth checking against the options, since 3 o'clock itself is a boundary the question may or may not include.

Example 6. Data sufficiency. What is the value of the integer ? Statement I: is a multiple of 6. Statement II: is a two-digit number whose digits sum to 9 and which is even.

Evaluate statement I alone. Multiples of 6 include 6, 12, 18 and infinitely many others, so it is insufficient.

Now forget statement I entirely and evaluate statement II alone. Two-digit numbers with digit sum 9 are 18, 27, 36, 45, 54, 63, 72, 81 and 90. Keeping only the even ones leaves 18, 36, 54, 72 and 90. Five candidates remain, so statement II alone is insufficient.

Together: the numbers surviving statement II are 18, 36, 54, 72 and 90, and every one of them is already a multiple of 6, since 18 = 6 times 3, 36 = 6 times 6, 54 = 6 times 9, 72 = 6 times 12 and 90 = 6 times 15.

So statement I adds nothing, and even together the two statements leave five possibilities. The answer is that the data are insufficient even when combined.

The instructive part is that combining statements does not always narrow anything. A candidate who assumed that two constraints must be more restrictive than one would have chosen the wrong option without checking.

Summary

Every analytical question gives constraints and asks what survives all of them, so choose the representation before reasoning.

Deduction preserves certainty and asks what must follow; induction preserves likelihood and asks what is best supported. Validity and truth are independent.

"All A are B" converts only to "Some B are A". Two negative or two particular premises yield nothing. Find the arrangement that breaks a conclusion, not one that supports it.

An assumption is tested by negation, a conclusion must be derivable from the statement alone, and a course of action must address the stated problem practically.

In critical reasoning, name the gap between evidence and conclusion; strengtheners close it and weakeners widen it.

For arrangements, draw the frame, place the most restrictive constraint first, and write the facing convention down before using left and right in a circle.

For blood relations, draw the tree and work outward one step at a time, reading words like "only" literally.

For directions, draw each leg as a vector, remember that turns are relative to the current facing, and recognise the standard Pythagorean triples.

The clock-hand angle is , and calendars reduce to odd days with 1 for an ordinary year and 2 for a leap year.

In data sufficiency, decide whether the answer is determined rather than computing it, and evaluate each statement in complete isolation.

Key formulas & results

Everything to memorise for the exam hall, in one card. Screenshot this for revision.

The organising tool
EVERY ANALYTICAL QUESTION GIVES CONSTRAINTS AND ASKS WHAT SURVIVES ALL OF THEM. NOTHING IS DISCOVERED BY INSIGHT; THINGS ARE ELIMINATED BY RULES.
THE DIFFICULTY IS ALMOST ENTIRELY REPRESENTATION. CHOOSE THE DIAGRAM FIRST, PLACE THE MOST RESTRICTIVE CONSTRAINT FIRST, AND LET THE REST ELIMINATE.
Validity is not plausibility
A CONCLUSION FOLLOWS ONLY IF IT MUST BE TRUE WHENEVER THE PREMISES ARE TRUE.
WHETHER IT IS TRUE IN THE WORLD, SENSIBLE, OR WHAT A REASONABLE PERSON WOULD DO IS IRRELEVANT, AND IS EXACTLY WHAT THE WRONG OPTIONS APPEAL TO.
Deduction versus induction
DEDUCTION GOES GENERAL TO PARTICULAR AND PRESERVES CERTAINTY. INDUCTION GOES PARTICULAR TO GENERAL AND PRESERVES ONLY LIKELIHOOD.
A DEDUCTION QUESTION ASKS WHAT MUST FOLLOW; AN INFERENCE QUESTION ASKS WHAT IS BEST SUPPORTED. AN OPTION THAT IS MERELY REASONABLE LOSES IN THE FIRST CASE.
Validity and truth are independent
ALL BIRDS CAN FLY; A PENGUIN IS A BIRD; THEREFORE A PENGUIN CAN FLY IS A VALID ARGUMENT WITH A FALSE CONCLUSION.
QUESTIONS EXPLOIT THIS BY SUPPLYING PREMISES THAT ARE OBVIOUSLY FALSE IN REALITY, AND CANDIDATES REJECT VALID CONCLUSIONS FOR BEING UNTRUE.
Valid conversions
ALL A ARE B GIVES SOME B ARE A. NO A IS B GIVES NO B IS A. SOME A ARE B GIVES SOME B ARE A. SOME A ARE NOT B GIVES NOTHING.
ALL A ARE B NEVER GIVES ALL B ARE A. THIS SINGLE CONFUSION ACCOUNTS FOR MORE ERRORS HERE THAN EVERYTHING ELSE COMBINED.
When nothing follows
TWO NEGATIVE PREMISES YIELD NO CONCLUSION. TWO PARTICULAR PREMISES, BOTH BEGINNING WITH SOME, YIELD NO CONCLUSION.
NEGATIVES ONLY EXCLUDE AND NEVER ESTABLISH THAT ANYTHING EXISTS IN A SHARED REGION; TWO SEPARATE OVERLAPS NEED NOT TOUCH EACH OTHER.
The counter-arrangement method
DRAW THE PREMISES AS REGIONS, THEN PUSH THE CIRCLES AS FAR APART AS THE PREMISES ALLOW. A CONCLUSION FOLLOWS ONLY IF IT SURVIVES EVERY PERMITTED ARRANGEMENT.
HUNT FOR THE ARRANGEMENT THAT BREAKS A CANDIDATE CONCLUSION RATHER THAN ONE THAT SUPPORTS IT. A SINGLE COUNTER-ARRANGEMENT SETTLES IT.
Either-or conclusions
AN EITHER-OR PAIR IS VALID WHEN THE TWO OPTIONS EXHAUST THE POSSIBILITIES AND NEITHER ALONE IS CERTAIN.
THIS IS THE ONE PLACE WHERE TWO INDIVIDUALLY INVALID CONCLUSIONS COMBINE INTO A VALID PAIR, TYPICALLY WHEN THE PREMISES LEAVE EXACTLY TWO ARRANGEMENTS.
The negation test for assumptions
IF DENYING THE CANDIDATE ASSUMPTION DESTROYS THE ARGUMENT, IT IS AN ASSUMPTION. IF THE ARGUMENT SURVIVES, IT IS NOT.
AN ASSUMPTION IS SOMETHING THE ARGUMENT TAKES FOR GRANTED AND NEEDS IN ORDER TO WORK, NOT MERELY SOMETHING CONSISTENT WITH IT.
Conclusion versus course of action
A CONCLUSION MUST BE DERIVABLE FROM THE STATEMENT ALONE. A COURSE OF ACTION MUST ADDRESS THE STATED PROBLEM AND BE PRACTICALLY IMPLEMENTABLE.
COURSES OF ACTION THAT ARE DRASTIC, THAT ADDRESS A DIFFERENT PROBLEM, OR THAT ASSUME AUTHORITY NOBODY HAS, FAIL EVEN WHEN THEY WOULD WORK.
Strengthen and weaken
LOCATE THE CONCLUSION, THEN THE EVIDENCE, THEN THE GAP BETWEEN THEM. EVERY STRENGTHENER CLOSES THAT GAP AND EVERY WEAKENER WIDENS IT.
AN OPTION THAT MERELY RESTATES THE CONCLUSION DOES NOT STRENGTHEN IT, AND ONE THAT ATTACKS A PREMISE'S PHRASING RATHER THAN THE INFERENCE DOES NOT WEAKEN IT.
The standard reasoning gaps
CORRELATION TREATED AS CAUSATION, UNREPRESENTATIVE SAMPLE, PART-TO-WHOLE LEAP, MISSING COMPARISON, PERCENTAGE VERSUS ABSOLUTE.
THE TYPICAL WEAKENERS ARE: AN ALTERNATIVE CAUSE, THE SAMPLE DIFFERS FROM THE POPULATION, THE CONTROL GROUP DID AS WELL, AND THE BASE CHANGED.
Affirming the consequent
FROM IF P THEN Q, AND Q, NOTHING FOLLOWS ABOUT P.
FROM IF THE PROGRAM COMPILES THEN THE SYNTAX IS CORRECT, PLUS THE SYNTAX IS CORRECT, NOTHING FOLLOWS ABOUT COMPILATION. DENYING THE ANTECEDENT IS THE MIRROR FALLACY.
Circular arrangement convention
FACING THE CENTRE, LEFT AND RIGHT ARE REVERSED RELATIVE TO HOW THEY LOOK ON THE PAGE. FACING OUTWARD, THEY READ NORMALLY.
HALF OF ALL CIRCULAR-ARRANGEMENT ERRORS COME FROM THIS ALONE. WRITE THE DIRECTION CONVENTION ON THE DIAGRAM BEFORE STARTING.
Between versus immediately between
BETWEEN IS AN ORDERING CONSTRAINT, NOT AN ADJACENCY, UNLESS THE WORD IMMEDIATELY OR EXACTLY APPEARS.
A CONSTRAINT OF THE FORM A IS NOT ADJACENT TO B IS OFTEN MORE POWERFUL THAN A POSITIVE ONE, SO APPLY NEGATIVE CONSTRAINTS EARLY RATHER THAN LAST.
Counting positions in an ordering
THIRD FROM THE TOP IN A GROUP OF NINE IS SEVENTH FROM THE BOTTOM. POSITION FROM ONE END PLUS POSITION FROM THE OTHER EQUALS TOTAL PLUS ONE.
THE OFF-BY-ONE HERE IS A STANDARD TRAP, AND THE FIX IS TO WRITE THE FULL LINE OF POSITIONS RATHER THAN COMPUTING MENTALLY.
Blood relation method
WORK STRICTLY FROM THE SPEAKER OUTWARDS, ONE RELATION AT A TIME, DRAWING A TREE WITH FIXED SYMBOLS FOR GENDER, MARRIAGE AND DESCENT.
GENDER IS OFTEN UNSTATED AND UNKNOWABLE, AND TERMS LIKE BROTHER-IN-LAW HAVE SEVERAL MEANINGS, SO SUCH QUESTIONS MAY ADMIT MORE THAN ONE TREE.
Directions
LEFT AND RIGHT TURNS ARE RELATIVE TO THE DIRECTION CURRENTLY FACED. A PERSON FACING SOUTH WHO TURNS LEFT FACES EAST.
COMPUTE NET VERTICAL AND NET HORIZONTAL DISPLACEMENT SEPARATELY AND COMBINE BY PYTHAGORAS. RECOGNISE 3-4-5, 5-12-13 AND 8-15-17 ON SIGHT.
Shadows
THE SUN RISES IN THE EAST AND SETS IN THE WEST, SO A MORNING SHADOW FALLS TOWARDS THE WEST AND AN EVENING SHADOW TOWARDS THE EAST.
AT NOON THE SHADOW IS SHORTEST AND ITS DIRECTION DEPENDS ON HEMISPHERE AND SEASON, WHICH IS WHY QUESTIONS STAY AWAY FROM NOON.
Letter coding
THE COMPLEMENT OF A LETTER'S POSITION IS 27 MINUS ITS POSITION, SO A PAIRS WITH Z, B WITH Y, AND M WITH N.
WRITE PLAINTEXT AND CODE ONE ABOVE THE OTHER, COMPUTE THE SHIFT PER LETTER, AND CHECK WHETHER IT IS CONSTANT, ALTERNATING OR INCREASING.
Testing a coding rule
A RULE THAT FITS TWO EXAMPLES AND FAILS A THIRD IS NOT THE RULE. TEST THE MAPPING ON EVERY GIVEN PAIR BEFORE APPLYING IT TO THE TARGET.
FOR DEFINED OPERATORS SUCH AS A STAR B MEANS A SQUARED MINUS B, SUBSTITUTE DIRECTLY AND WORK FROM THE INNERMOST BRACKET OUTWARDS.
Clock hand angle
THE ANGLE BETWEEN THE HANDS AT h HOURS AND m MINUTES IS THE ABSOLUTE VALUE OF 30h MINUS 5.5m, TAKING THE SMALLER OF THAT AND 360 MINUS IT.
THE HOUR HAND MOVES 0.5 DEGREES PER MINUTE AND THE MINUTE HAND 6, SO THE MINUTE HAND GAINS 5.5 DEGREES PER MINUTE.
Clock coincidences
THE HANDS COINCIDE 11 TIMES IN 12 HOURS AND ARE AT RIGHT ANGLES 22 TIMES IN 12 HOURS.
IT IS 11 AND NOT 12 BECAUSE BETWEEN 11 O'CLOCK AND 1 O'CLOCK THEY COINCIDE ONLY ONCE, AT 12.
Calendars and odd days
AN ORDINARY YEAR HAS 1 ODD DAY AND A LEAP YEAR HAS 2. A CENTURY HAS 5 ODD DAYS AND 400 YEARS HAS 0.
THE CALENDAR THEREFORE REPEATS EXACTLY EVERY 400 YEARS. ODD DAYS ARE THE REMAINDER WHEN A DAY COUNT IS DIVIDED BY 7.
The leap-year exception
DIVISIBLE BY 4, EXCEPT CENTURIES, UNLESS DIVISIBLE BY 400. SO 1900 WAS NOT A LEAP YEAR AND 2000 WAS.
QUESTIONS ARE CONSTRUCTED SPECIFICALLY AROUND THIS EXCEPTION, SO A CENTURY YEAR IN THE STEM IS ALWAYS DELIBERATE.
Data sufficiency discipline
DECIDE WHETHER THE ANSWER IS DETERMINED, DO NOT COMPUTE IT. EVALUATE STATEMENT I COMPLETELY, THEN DELIBERATELY FORGET IT BEFORE EVALUATING STATEMENT II.
A STATEMENT IS SUFFICIENT WHEN IT FORCES A UNIQUE ANSWER. FINDING TWO VALUES CONSISTENT WITH IT IS THE FASTEST PROOF OF INSUFFICIENCY.
⚠️

Traps GATE sets — and how to dodge them

These are the exact option-traps and misreads that cost marks under negative marking.

WATCH OUT
Rejecting a valid conclusion because the premises are false in reality
Validity and truth are independent. If the premises say all birds can fly, then within the question a penguin can fly. The task is to check whether the conclusion follows, not whether it is true.
WATCH OUT
Converting 'All A are B' to 'All B are A'
The only valid conversion is 'Some B are A'. All engineers are graduates does not make all graduates engineers, and this is the single largest source of syllogism errors.
WATCH OUT
Drawing a conclusion from two particular or two negative premises
Both combinations yield nothing. Recognising this eliminates every definite-conclusion option before any diagram is drawn, which is usually enough to answer the question.
WATCH OUT
Looking for an arrangement that supports the conclusion
Look for one that breaks it. A conclusion follows only if it survives every arrangement the premises permit, so a single counter-arrangement is decisive while a supporting one proves nothing.
WATCH OUT
Treating a plausible statement as an assumption
Apply the negation test. If denying it leaves the argument intact, it is not an assumption but merely a compatible fact. An assumption is something the argument needs in order to work.
WATCH OUT
Choosing a course of action because it would solve the problem
It must also address the problem as stated and be practically implementable by whoever is acting. Drastic measures, and actions assuming authority nobody has, fail even when they would work.
WATCH OUT
Selecting a strengthener that restates the conclusion
A restatement adds no support. Name the gap between the evidence and the conclusion first; the correct strengthener closes that specific gap and the correct weakener widens it.
WATCH OUT
Affirming the consequent
From 'if P then Q' and Q, nothing follows about P. The syntax being correct does not mean the program compiled, because other things can also make syntax correct. Denying the antecedent is the mirror error.
WATCH OUT
Reasoning through a seating puzzle without drawing it
Draw the frame with numbered positions first. The puzzle is a constraint-satisfaction problem, and the constraints only eliminate efficiently once there is something for them to eliminate from.
WATCH OUT
Reading left and right normally in a circular arrangement
Facing the centre reverses them relative to the page. Write the facing convention on the diagram before placing anyone, since half of all circular-arrangement errors come from this alone.
WATCH OUT
Reading 'between' as 'immediately between'
Unless the question says immediately or exactly, 'between' only fixes an ordering. Treating it as adjacency over-constrains the puzzle and typically makes it appear to have no solution.
WATCH OUT
Applying negative constraints last
A constraint of the form 'A is not adjacent to B' often eliminates several arrangements at once. Applying it early rather than as a final check saves substantial time in a timed section.
WATCH OUT
Miscounting positions from the two ends
Position from one end plus position from the other equals the total plus one. Third from the top in a group of nine is seventh, not sixth, from the bottom. Write the full line rather than computing mentally.
WATCH OUT
Skipping over qualifying words in a blood-relation stem
Words like 'only' change the answer completely. The only son of my grandmother is my father; without 'only' he could be an uncle, making the woman a cousin rather than a sister.
WATCH OUT
Tracking direction turns mentally
Draw each leg. Turns are relative to the direction currently faced, so a person facing east who turns right faces south. Tracking distances and facings simultaneously in the head is where errors enter.
WATCH OUT
Accepting a coding rule that fits only the first example
Test the mapping against every given pair before applying it. A shift that fits two letters and fails the third is not the rule, and the failure usually reveals an alternating or increasing pattern.
WATCH OUT
Forgetting the century exception to the leap-year rule
Divisible by 4, except centuries, unless divisible by 400. A century year appearing in a calendar question is always deliberate, since 1900 was not a leap year while 2000 was.
WATCH OUT
Carrying statement I into the evaluation of statement II
Evaluate each statement in complete isolation, then together. Carrying information across is the commonest data-sufficiency error and produces the wrong answer in exactly the cases the question was built to test.
WATCH OUT
Assuming two statements together are always more restrictive than one
Not always. If every value satisfying statement II already satisfies statement I, then combining them adds nothing and the data can remain insufficient even together.

Exam-pattern practice

PYQ-style questions with full solutions. Work through them as a readiness check — mark yourself honestly and get your gap report at the end.

Readiness check

Are you exam-ready for Analytical Aptitude?

9 problems from this chapter. Try each one, reveal the worked solution, mark yourself honestly — get your gap report at the end.

9 questions~6 min worth ~100 marks in GATE exams

5-minute revision

The whole chapter, distilled. Read this the night before the exam.

  • Every analytical question is constraints plus elimination.
  • Choose the representation before reasoning.
  • Place the most restrictive constraint first.
  • Validity is not plausibility.
  • Deduction preserves certainty; induction preserves likelihood.
  • A valid argument can have a false conclusion if a premise is false.
  • All A are B converts only to Some B are A.
  • No A is B converts to No B is A.
  • Some A are not B converts to nothing.
  • Two negative premises yield no conclusion.
  • Two particular premises yield no conclusion.
  • Hunt for the arrangement that breaks a conclusion.
  • Either-or pairs are valid when the options exhaust the possibilities.
  • Test an assumption by negating it.
  • A conclusion must come from the statement alone.
  • A course of action must be practical and address the stated problem.
  • Name the gap between evidence and conclusion first.
  • Correlation treated as causation is the commonest gap.
  • A restatement of the conclusion does not strengthen it.
  • Affirming the consequent proves nothing about the antecedent.
  • Draw the frame before placing anyone in an arrangement puzzle.
  • Facing the centre reverses left and right.
  • Between is not immediately between unless stated.
  • Negative adjacency constraints eliminate fastest.
  • Position from one end plus from the other equals total plus one.
  • Work blood relations outward from the speaker.
  • Read qualifying words like 'only' literally.
  • Gender is often unstated and may be unknowable.
  • Turns are relative to the current facing.
  • Compute net vertical and horizontal displacement separately.
  • Recognise 3-4-5, 5-12-13 and 8-15-17 on sight.
  • Morning shadows fall west, evening shadows east.
  • A letter's complement position is 27 minus its position.
  • Test a coding rule on every given pair.
  • The clock angle is 30h minus 5.5m in absolute value.
  • The minute hand gains 5.5 degrees per minute.
  • The hands coincide 11 times in 12 hours.
  • An ordinary year has 1 odd day, a leap year 2.
  • A century has 5 odd days; 400 years has 0.
  • Divisible by 4, except centuries, unless divisible by 400.
  • In data sufficiency, decide rather than compute.
  • Evaluate each statement in complete isolation.
  • Two values consistent with a statement prove it insufficient.
  • Combining statements does not always narrow anything.

GATE question blueprint

How this topic is asked, tier by tier — so you can prep to the pattern.

Typical weightage: General Aptitude is 15 of the 100 marks in every GATE paper: 5 questions at 1 mark and 5 at 2 marks. Analytical reasoning typically supplies 2-3 of those 10 questions

Question styleMarks eachTypical countWhat it tests
Syllogisms1~1Valid conversions, premise combinations that yield nothing, and the counter-arrangement method
Critical reasoning2~1Locating the gap between evidence and conclusion, and choosing a genuine strengthener or weakener
Arrangements2~1Constraint ordering, circular facing conventions and adjacency versus ordering
Blood relations1~1Working outward from the speaker and reading qualifying words literally
Directions1~1Relative turns, net displacement and the standard Pythagorean triples
Coding and numerical relations2~1Positional shifts, testing a rule against every pair, and defined operators
Clocks and calendars2~1The hand-angle formula, odd days and the century exception to the leap-year rule
Data sufficiency2~1Deciding rather than computing, and evaluating statements in isolation

Exam-hall strategy

Battle-tested tips from mentors and toppers for this topic under the sectional clock.

  1. Draw the representation before reading the question again.
  2. Place absolute and negative constraints before relative ones.
  3. For syllogisms, check for two negatives or two particulars before drawing anything.
  4. Test any candidate conclusion by trying to break it, not to support it.
  5. For assumptions, negate the option and see whether the argument survives.
  6. In data sufficiency, decide sufficiency and never compute the answer.
  7. GA analytical items are a mix of 1-mark and 2-mark MCQs with -1/3 and -2/3 for wrong answers, so guess only after eliminating an option.
  8. If an analytical item is set as an MSQ, attempt it regardless of confidence, since MSQ carries no negative marking.
  9. GATE gives a single freely-navigable 180-minute window, so flag a long constraint puzzle and return to it once the technical sections are secured.

Beyond the exam

Where this skill shows up in the job you're competing for — and in life.

Debugging by elimination

Listing the constraints a failing system must satisfy and eliminating hypotheses that violate them is the same procedure as solving a seating puzzle, and it is faster than guessing.

Reviewing a causal claim in a postmortem

Recognising that a correlation between a deploy and an outage needs an alternative-cause check is exactly the weaken-the-argument move in critical reasoning.

Deciding whether requirements are sufficient

Data sufficiency is the everyday question of whether a specification determines a unique behaviour, and the method — find two systems satisfying it — is identical.

Reading a scheduling constraint precisely

Distinguishing 'between' from 'immediately between' is the same literal reading that prevents a scheduling or dependency requirement from being implemented too strictly.

Where else this topic is tested

Prepare once, score in every exam that asks it.

GATE DA and other GATE papersIdentical — General Aptitude is a common 15-mark section across every GATE paper, with the same question shapes
CAT / XAT logical reasoningHigh overlap — arrangement puzzles, critical reasoning and data sufficiency are examined in the same form at greater length and difficulty
UGC NET Paper 1 and SSC CGL reasoningHigh overlap — syllogisms, blood relations, coding-decoding, directions and calendars are shared almost topic for topic

Questions aspirants ask

Pulled from the Q&A community and mentor sessions.

Because validity is a property of the argument's form and truth is a property of its content, and the two are independent. An argument is valid when it is impossible for the premises to be true and the conclusion false. That says nothing about whether the premises actually are true. 'All birds can fly; a penguin is a bird; therefore a penguin can fly' is a perfectly valid argument whose conclusion is false, and it is false only because the first premise is false. GATE exploits this deliberately. Questions supply premises such as 'All engineers are poets' precisely so that candidates who evaluate by real-world plausibility get the wrong answer. Within the question, the premises are stipulated true, and the only task is to check whether the conclusion is forced. The reverse trap also appears: a conclusion that is obviously true in the world but not entailed by the premises. 'Some engineers are researchers' is true in reality, and a syllogism question can present it as a conclusion that does not follow, which catches candidates who accept it because they know it to be so. The reliable habit is to treat the premises as a small closed world, draw them, and ask only whether the conclusion survives every arrangement that world permits.

It is the observation that a conclusion follows only if it holds in every arrangement the premises permit, so a single arrangement in which it fails settles the matter immediately. Checking that a conclusion holds requires examining all arrangements; checking that it fails requires finding one. The asymmetry is what makes the method fast. In practice, draw the premises as regions and then deliberately push the circles as far apart as the premises still allow. If the premise is 'All A are B', A must sit inside B, but B can be much larger with room to spare, and any third region can be placed in that spare room. If the premise is 'Some A are B', the overlap can be as small as a single element, and everything else can be pushed outside. Applying this to 'All engineers are graduates; some graduates are researchers' and the candidate conclusion 'Some engineers are researchers': place the researcher region so that it overlaps graduates entirely outside the engineer region. Both premises hold, the conclusion fails, and the question is answered without considering any other arrangement. The habit worth building is to construct the most spread-out arrangement first, since that is where conclusions break, rather than the neat overlapping picture that makes everything look true.

Each has its own test, and applying the wrong test is what makes these three feel interchangeable. An assumption is something the argument needs but does not state. The test is negation: deny the candidate and see whether the argument collapses. If a college installs cameras to deter misconduct, the assumption is that cameras deter or record misconduct; deny that, and the decision makes no sense, so it is an assumption. Something merely compatible with the argument, such as 'cameras are affordable', survives negation without destroying the reasoning and is therefore not an assumption. A conclusion must be derivable from the statement alone, with no added information. From the same statement, 'the college has some concern about laboratory conduct' follows, because installing cameras there implies a concern. 'The misconduct will stop' does not follow, because the statement reports a decision, not an outcome. A course of action is judged on two things at once: does it address the problem as stated, and can whoever is acting actually implement it. Briefing students on the new policy qualifies. Expelling all students found in a laboratory after hours does not, because it is disproportionate and addresses a different problem. The commonest error is choosing a course of action because it would work, without checking that it targets the stated problem and lies within the actor's authority.

Because 'left' and 'right' change meaning depending on which way the people face, and the change is counter-intuitive on paper. If everyone faces the centre of the circle, then from their point of view, someone to their left appears on the clockwise side when you look at the diagram from outside. If everyone faces outward, the mapping is the reverse. Half of all errors in these puzzles come from this single ambiguity, and none of them come from the logic. The fix is procedural rather than conceptual: before placing anybody, write on the diagram which way people face and mark one position with an arrow showing which direction counts as that person's left. Every subsequent constraint is then applied against a fixed reference rather than being re-derived each time. A second, smaller source of error is mixed facing, where the question states that some people face inward and others outward. Here the left-right mapping differs per person, and the only safe approach is to annotate each seat individually as it is filled. Finally, remember that in a circular arrangement of n people, rotations are the same arrangement, so 'A is third to the left of B' is a relative constraint and does not fix anyone's absolute seat. Absolute positions only exist when the question supplies a labelled seat or a fixed reference point.

The question asks whether the answer is determined, not what the answer is, and those are different tasks with different costs. Deciding sufficiency often takes one observation: if you can produce two different values consistent with a statement, it is insufficient, and you are done. Computing the actual answer requires carrying the arithmetic to completion, which takes longer and produces no additional information. The bigger cost is subtler. While computing under statement I, candidates naturally use everything they have read, including statement II, and the contamination is almost impossible to notice from the inside. The result is judging statement I sufficient when it is not, which is precisely the error the question format was designed to catch. The discipline is mechanical. Evaluate statement I in isolation and record a verdict. Then deliberately treat statement II as if statement I had never been read, evaluate it, and record a second verdict. Only then consider them together. It is also worth remembering that combining does not always help. If every value satisfying statement II already satisfies statement I, then statement I adds nothing and the pair remains insufficient. And two statements can be mutually inconsistent, in which case nothing satisfies both and no answer exists at all — a case that only appears if consistency is checked rather than assumed.
Header Logo