Module 3 · SSC GD Reasoning

Analogy

Updated 17 Nov 2026 · 6,438 words · illustrated
Updated 17 Nov 2026
Word, letter, and number analogies.
Word Analogies · Letter Analogies · Number Analogies
🎯 Take Mock Test
Identify SeriesDetect PatternApply RuleSelect AnswerCommon PatternsTransformationTypesnextnextchooseArt. 1: Analogy = Similar RelationMechanism: Same operation on each termvlymbooq
Series Analogy Reasoning Steps for Geography Exam Preparation — The diagram maps the four-step reasoning chain used to solve series analogy questions in GS‑1 Geography. · Open the full diagram →
Structural AnalogyProcess AnalogyFunctional AnalogyLandform ComparisonClimateClassificationHazard Assessmentapplies toguidesinformsAnalogy aids spatial reasoning (Art. 1)Key for UPSC GS1 map interpretationvlymbooq
Analogy Types and Applications in Physical Geography for UPSC GS1 — The diagram classifies geographic analogies and links each type to its practical use in map interpretation. · Open the full diagram →

Learning Objectives

  • Understand the fundamental concepts of Analogy
  • Apply key formulas and techniques to solve problems
  • Practice with exam-level questions to build speed and accuracy

What is Analogy?

Analogy means 'similarity' or 'correspondence'. In reasoning, an analogy question presents a pair of items that share a specific relationship. You must identify the same relationship in another pair and find the missing term. These questions test your ability to recognize patterns and relationships between words, letters, or numbers.

Types of Analogies

1. Word Analogies: Based on semantic relationships between words — synonym, antonym, cause-effect, part-whole, worker-tool, product-raw material, animal-young one, etc. Example: Doctor : Hospital :: Teacher : School (worker-workplace relationship).

2. Letter Analogies: Based on positional values of letters (A=1, B=2, ..., Z=26) or alphabetical patterns. Example: ACE : FHJ :: MOQ : PRT (each letter advances by 5 positions).

3. Number Analogies: Based on arithmetic operations between numbers — addition, subtraction, multiplication, division, squares, cubes, or a combination. Example: 4 : 20 :: 6 : 42 (4×5=20, 6×7=42 — multiply by consecutive odd numbers starting from 5).

Solved Examples

Example 1 (Word Analogy)
Find the missing term: Puppy : Dog :: Kitten : ?

Options: a) Cow b) Cat c) Horse d) Sheep
Solution: b) Cat. Puppy is a young dog. Similarly, Kitten is a young Cat. This is an animal-young one relationship.
Example 2 (Letter Analogy)
Find the missing term: BEH : KNQ :: ? : TXZ

a) MOP b) OQS c) PSU d) QTW
Solution: c) PSU. In BEH, B→E (+3), E→H (+3). In KNQ, K→N (+3), N→Q (+3). So pattern is +3 each step. Working backwards from TXZ: T←Q (-3), Q←N (-3) = QN...
Example 3 (Number Analogy)
Find the missing term: 3 : 27 :: 4 : ?

Options: a) 64 b) 16 c) 81 d) 256
Solution: a) 64. 3³ = 27. Similarly, 4³ = 64. This is a square/cube relationship.

Shortcut Techniques

  • Always identify the exact relationship in the first pair before looking at options.
  • For word analogies, think: category, function, part-whole, cause-effect, degree, synonym-antonym.
  • For letter analogies, convert letters to numbers (A=1 to Z=26) and look for arithmetic patterns.
  • For number analogies, test basic operations in order: ×, ÷, +, −, then squares/cubes.
Pro Tip
When stuck between two options, identify the secondary relationship. For example, Doctor: Hospital involves both workplace AND profession. Make sure both dimensions match.

Advanced Problem Types

Competitive exams feature advanced variations combining multiple reasoning concepts. Master these to score high.

Multi-Step Problems

Some questions require applying multiple reasoning techniques sequentially. Break problems into clear steps.

Common Traps

  • Assuming a pattern continues without checking all elements
  • Misinterpreting negative statements in syllogisms
  • Forgetting wrapping in alphabet coding (Z to A transition)
  • Confusing directions (East vs West)
  • Overlooking either-or cases in syllogisms
  • Not considering all possible arrangements in puzzles
Pro Tip
For puzzles, start with the most concrete clue and build around it. Draw diagrams for seating arrangements and family trees for blood relations.

Compound Analogies Combining Word and Number Logic

In many high‑difficulty items the first pair mixes a semantic relationship with a numeric operation, and the second pair demands the same dual logic. The examiner expects the candidate to isolate two independent rules — one governing the words, the other governing the numbers — and then apply both simultaneously. A typical structure is Word₁ : Number₁ :: Word₂ : Number₂ where the word pair follows a category‑member or function‑tool relation while the number pair follows a mathematical transformation such as “multiply by the position of the first letter in the alphabet”.

To solve, first map the word relationship. If Doctor : Stethoscope appears, the link is “profession : instrument”. Next, examine the numeric side. Suppose the given numbers are 4 : 20. Test elementary operations: 4 × 5 = 20, 4² + 4 = 20, 4 + 16 = 20. The simplest consistent rule is “multiply by the next integer (5)”. Verify whether the same multiplier works for the second word’s numeric counterpart. If the second word is Carpenter and the options include 30, 36, 42, 48, compute the position of the first letter C = 3; 3 × ? must equal the answer. The multiplier that fits the first pair (5) suggests the rule “multiply the alphabetical index of the initial letter by 5”. Thus 3 × 5 = 15, but 15 is not an option — indicating the multiplier may be derived from the word length (Doctor = 6 letters → 6 + ? = 5? No). The correct approach is to treat the two relations independently: the word pair tells you the category, the number pair tells you the arithmetic. In the example, the word pair is only a distractor; the numeric rule is “n × (n + 1)”. Hence 4 × 5 = 20 and 6 × 7 = 42. The answer is the option matching 42.

Key checkpoints: (1) Write down the word relation in plain language. (2) List all plausible arithmetic rules for the first numeric pair. (3) Test each rule on the second numeric slot using the data derived from the second word (letter position, word length, vowel count, etc.). (4) Discard rules that need external information not present in the question. (5) Choose the rule that yields an answer present in the options.

Common variations include: (a) Word pair shows synonym/antonym while numbers follow square‑cube pattern; (b) Word pair shows part‑whole while numbers follow prime‑gap sequence; (c) Word pair shows cause‑effect while numbers follow alternating addition/subtraction. Mastery comes from practising a matrix of word‑type × number‑type combinations until the separation of concerns becomes automatic.

Positional Letter Patterns with Wrap‑Around and Reverse Sequences

Letter analogies frequently exploit the cyclic nature of the alphabet. A forward shift of +3 from Y yields B (Y→Z→A→B). Reverse shifts behave similarly: a –4 shift from C yields Y (C→B→A→Z→Y). Questions may combine forward and reverse steps within a single triplet, e.g., ACF : PRU :: GJM : ?. The first triplet shows positions 1, 3, 6; the second shows 16, 18, 21. The differences are +15, +15, +15 — a uniform Caesar shift of 15. Applying the same shift to GJM (7, 10, 13) gives 22, 25, 28 → 28 wraps to 2 (B), producing V, Y, B → VYB.

More intricate patterns alternate direction: +2, –3, +2, –3 … or use a keyword to determine shift values (Vigenère‑style). For instance, BDF : KMO :: ? : SUW. BDF (2, 4, 6) → KMO (11, 13, 15) shows +9 each. The reverse pair demands a triplet that, when shifted +9, yields SUW (19, 21, 23). Subtract 9: 10, 12, 14 → J, L, N → JLN. When the shift is not constant, compute the difference series for each position separately.

Wrap‑around traps appear when the shift pushes a letter past Z or before A. Always convert letters to 1‑26 numbers, apply modular arithmetic (mod 26), then convert back. A shift of +27 is equivalent to +1; a shift of –27 equals –1. In timed conditions, memorise the 26‑letter circle and practice mental modulo: (position + shift – 1) mod 26 + 1.

Reverse‑sequence analogies present the second pair in descending order, e.g., XYZ : CBA :: MNO : ?. XYZ (24, 25, 26) → CBA (3, 2, 1) is a mirror about the centre (27 – position). The rule: each letter maps to its “alphabetic opposite” (A↔Z, B↔Y, …). Applying to MNO (13, 14, 15) gives 14, 13, 12 → N, M, L → NML. Recognise mirror, reverse, and rotational patterns by checking whether the sum of paired positions equals 27.

Examiners also embed a “skip‑one” pattern: positions 1, 3, 5, 7 … or 2, 5, 8 … (arithmetic progression with common difference 3). Write the numeric positions, compute differences, and test whether the same progression holds for the target triplet.

Worked Example – Mixed Shift and Mirror
Find the missing term: FHK : UQM :: JLP : ? Options: a) QJO b) QJN c) QKM d) QKN

Step 1: Convert letters to numbers (A=1). FHK → 6, 8, 11. UQM → 21, 17, 13.

Step 2: Compute position‑wise differences: 21‑6 = +15, 17‑8 = +9, 13‑11 = +2. The differences are not uniform, so look for a secondary pattern. Observe that 15, 9, 2 are decreasing by 6 then 7 — not a simple arithmetic series.

Step 3: Test mirror hypothesis: 27 – 6 = 21 (U), 27 – 8 = 19 (S) but we have 17 (Q). So not pure mirror.

Step 4: Consider a keyword shift. The first letters F (6) → U (21) is +15. The second letters H (8) → Q (17) is +9. The third letters K (11) → M (13) is +2. The shifts 15, 9, 2 correspond to the positions of the letters in the word “FHK” themselves? 6, 8, 11 → no. But 15 = 6+9, 9 = 8+1, 2 = 11‑9 — not consistent.

Step 5: Re‑examine the triplet as a whole. Write the alphabet indices of the first triplet: 6, 8, 11. The second triplet: 21, 17, 13. Notice that 6+21 = 27, 8+17 = 25, 11+13 = 24. The sums are 27, 25, 24 — decreasing by 2 then 1. This suggests the rule: each pair sums to a descending sequence starting at 27. For the third pair, the sum should be 23 (24‑1).

Step 6: Apply to JLP → 10, 12, 16. Required partners: 27‑10 = 17 (Q), 25‑12 = 13 (M), 23‑16 = 7 (G). That yields QMG, not in options.

Step 7: Alternative simpler rule: The second triplet is the reverse of the first triplet after a Caesar shift of +15 on the first letter only, then –? Let's test the options. Option a) QJO → 17, 10, 15. Compare with JLP (10, 12, 16). Differences: +7, –2, –1. Option b) QJN → 17, 10, 14 → +7, –2, –2. Option c) QKM → 17, 11, 13 → +7, –1, –3. Option d) QKN → 17, 11, 14 → +7, –1, –2.

Step 8: The first difference +7 appears in all options, matching the shift from F (6) to U (21) which is +15, not +7. However, note that the first letters of the two given triplets are F (6) and U (21). Their average is (6+21)/2 = 13.5. The first letter of the third triplet J (10) and the answer’s first letter Q (17) average to 13.5 as well. This “mid‑point” property holds for the first position only.

Step 9: Check second position: H (8) and Q (17) average 12.5. L (12) and answer’s second letter should average 12.5 → answer second = 13 (M). Only options c) and d) have M as second letter.

Step 10: Third position: K (11) and M (13) average 12. P (16) and answer third should average 12 → answer third = 8 (H). None of the options have H. Therefore the midpoint rule fails for third position.

Step 11: Re‑evaluate using the most consistent pattern: first position shift +15, second position shift +9, third position shift +2. Apply same shifts to JLP: J (10)+15 = 25 (Y), L (12)+9 = 21 (U), P (16)+2 = 18 (R) → YUR, not an option.

Step 12: Since none of the systematic rules match the options, the intended pattern is likely the “average of first letters” and “average of second letters” only, with the third letter following a simple reverse order (K→M is +2, so P→? should be –2 → N). That gives QMN, not listed. The closest is QKN (option d) where third letter N is –2 from P (16→14). Second letter K (11) is –1 from L (12). First letter Q (17) is +7 from J (10). The shifts +7, –1, –2 appear in option d. The first pair shifts were +15, +9, +2. The second pair shifts are roughly half of the first pair (rounded). Hence the rule: each shift in the second analogy is half the corresponding shift in the first analogy (rounded to nearest integer). 15/2≈7.5→7, 9/2=4.5→5 (but we have –1), 2/2=1→1 (but we have –2). Not perfect.

Given the ambiguity, the most defensible answer based on the only consistent half‑shift for the first position is option d) QKN.

Answer: d) QKN

Exam‑Day Strategy for Mixed Letter‑Number Analogies
When a question presents a hybrid of word, letter, and number logic, allocate the first 15 seconds to label each component (word‑relation, letter‑pattern, number‑rule). Write the three rules in shorthand on the rough sheet. If any rule cannot be fixed within 20 seconds, flag the item and move on; such questions often have a single dominant pattern (usually the numeric one) that yields the answer directly. Return only if time permits.

Number Analogies Based on Prime, Fibonacci, and Polynomial Functions

Beyond squares and cubes, competitive papers increasingly use sequences derived from prime numbers, Fibonacci numbers, and low‑degree polynomials. A prime‑based analogy may read 2 : 3 :: 5 : ? where the rule is “next prime”. The answer is 7. A more concealed version: 4 : 7 :: 10 : ? where 4 and 10 are composite, but 7 is the 4th prime; the rule could be “nth prime where n is the first number”. Then 10th prime = 29.

Fibonacci analogies exploit the recurrence Fₙ = Fₙ₋₁ + Fₙ₋₂. Example: 3 : 5 :: 8 : ?. Here 3 and 5 are consecutive Fibonacci numbers; 8 is the next. The missing term is the Fibonacci number after 8, i.e., 13. Another variant: 5 : 34 :: 8 : ?. 5th Fibonacci = 5, 9th Fibonacci = 34 (since F₉ = 34). The rule: “first number indicates index, second gives the Fibonacci value at index + 4”. Then for 8, index + 4 = 12 → F₁₂ = 144.

Polynomial functions appear as “n² + n”, “n³ – n”, or “2n² + 3n + 1”. For 2 : 10 :: 5 : ?, test n² + n: 2²+2=6 (no). n³ + 2: 2³+2=10 (yes). Apply to 5: 5³+2=127. If 127 is an option, that is the answer. Always generate a short list of candidate polynomials (degree ≤ 3) using the first pair, then verify with the second pair’s first element.

Mixed prime‑polynomial analogies: 6 : 13 :: 15 : ?. 6th prime = 13. So rule: “first number → that‑th prime”. 15th prime = 47. If options include 47, select it. If not, consider “first number + 7 = prime” (6+7=13). Then 15+7=22 (not prime). Hence the index‑prime rule is stronger.

Examiners also combine arithmetic and geometric progressions: 3 : 12 :: 5 : ?. 3 × 4 = 12. 5 × ? = ?. The multiplier may be the first number + 1 (3+1=4). Then 5 × 6 = 30. Alternatively, multiplier could be the next prime after the first number (next prime after 3 is 5 → 3 × 5=15, not 12). So the “+1” rule fits. Always test the simplest linear multiplier first.

When numbers are large, factorise quickly: 72 = 2³ × 3², 108 = 2² × 3³. The analogy 72 : 108 :: 200 : ? may involve swapping exponents (2³ × 3² → 2² × 3³). Apply to 200 = 2³ × 5² → swap → 2² × 5³ = 4 × 125 = 500. Recognising prime‑factor patterns saves time.

Hybrid Letter‑Number Coding Analogies

These items encode letters as numbers (A=1 … Z=26) and then apply arithmetic or positional operations before converting back. A classic format: CODE : 3‑15‑4‑5 :: WORD : ?. The first pair shows the direct alphabetic positions. The second pair may require a transformation such as “add 2 to each position” → W(23)+2=25(Y), O(15)+2=17(Q), R(18)+2=20(T), D(4)+2=6(F) → YQTF. Options will present either the numeric string or the letter string.

More complex coding uses a keyword to vary the shift per position (Vigenère). Example: KEY : 11‑5‑25 :: LOCK : ? with keyword “KEY”. Convert KEY → 11, 5, 25. For LOCK (12, 15, 3, 11), add keyword cyclically: 12+11=23(W), 15+5=20(T), 3+25=28→2(B), 11+11=22(V) → WTBV. The answer may be given as letters or as numbers 23‑20‑2‑22.

Another variant: “multiply position by its index in the word”. For CAT : 3‑2‑20 :: DOG : ?, C(3)×1=3, A(1)×2=2, T(20)×3=60 → 3‑2‑60. For DOG: D(4)×1=4, O(15)×2=30, G(7)×3=21 → 4‑30‑21.

Reverse coding: given the numeric string, recover the original word. ? : 8‑5‑12‑12‑15 :: HELLO : 8‑5‑12‑12‑15. The first pair is identical, so the missing term is HELLO. But if the second pair shows a transformation (e.g., each number +1), then the first pair must be the pre‑transformation numbers. Practice both directions.

Hybrid analogies often combine a letter pattern (like +3, –2) with a numeric pattern (like ×2). Example: ACE : 2‑4‑6 :: GIK : ?. ACE positions 1, 3, 5 → doubled → 2, 6, 10 but given 2‑4‑6 suggests a different rule: “position of each letter in the word” (A=1st, C=2nd, E=3rd) → 1, 2, 3 then ×2 → 2, 4, 6. For GIK (positions 7, 9, 11) → word positions 1, 2, 3 → ×2 → 2, 4, 6 again. So answer is 2‑4‑6. Recognising “word‑index” versus “alphabet‑index” is crucial.

Multi‑Step Transformation Chains in Analogy Questions

Some questions require a chain of three or more transformations applied sequentially to the first term to reach the second term, and the same chain must be reproduced for the third term. For instance, 12 : 30 :: 18 : ? with hidden steps: (1) add 3 → 15, (2) multiply by 2 → 30. Apply to 18: 18+3=21, 21×2=42. The answer is 42. The chain may be described implicitly; the candidate must infer the minimal sequence that fits both given pairs.

Letter chains work similarly: ABC : DEF :: GHI : ?. Steps: each letter +3. But a multi‑step chain could be: reverse the triplet, then shift each letter +1. ABC → CBA → DCB. If the given second term is DCB, the chain is identified. Apply to GHI: reverse → IHG, shift +1 → JIH.

Word chains may involve semantic steps: Seed : Tree :: Egg : ?. Step 1: “grows into” (seed→plant). Step 2: “mature form” (plant→tree). For egg, step 1: “hatches into” (egg→chick). Step 2: “adult form” (chick→hen). Answer: Hen. The chain is “developmental stage → mature organism”. Recognising the two‑stage biological progression avoids the trap of answering “Chick”.

Numeric‑letter hybrid chains: 2B : 6F :: 3D : ?. Decode: number part 2→6 (×3), letter part B→F (+4). Apply to 3D: 3×3=9, D+4=H → 9H. If options include 9H, select it. Chains can have three operations: e.g., number: +2, ×2, –1; letter: +1, –2, +3. Write each operation as a function and compose.

To crack chains quickly, list all observable differences between the two given terms (numeric delta, letter shifts, word relations). Then test whether a single consistent sequence of elementary operations (addition, multiplication, reversal, synonym shift) reproduces the second term from the first. If multiple sequences fit, choose the one with the fewest steps (Occam’s razor). Practice with timed drills of 20‑30 chain items to internalise common operation sets.

  • Assuming the first obvious numeric operation (e.g., square) is the only rule without testing alternatives.
  • Ignoring wrap‑around in alphabetic shifts and treating Z→A as a break in the pattern.
  • Confusing word‑index (position inside the word) with alphabet‑index (A=1) in hybrid coding.
  • Applying a single‑step rule to a multi‑step chain and selecting an answer that matches only the first step.
  • Overlooking prime‑index or Fibonacci‑index relationships when numbers are small but not perfect squares or cubes.
  • Treating a reverse‑order letter pair as a simple Caesar shift instead of a mirror (sum = 27) pattern.
  • Failing to verify that the derived rule works for all given positions before committing to an option.

Consistent practice with these advanced patterns builds the pattern‑recognition speed required for high‑stakes reasoning sections.

Recursive Analogy Structures and Self‑Referential Patterns

Recursive analogies appear when the relationship between the first pair is itself defined by an analogy that mirrors the overall structure. In such items the rule is not a simple arithmetic or lexical transformation but a meta‑rule: “the way A relates to B is the same way the rule relates to its own description.” A typical format presents two pairs where the second pair encodes the rule used in the first pair. For example, AB : CD :: EF : ? might hide the rule “shift each letter forward by the position of the first letter in the alphabet.” If A=1, then shift=1, giving BC; but the given second term is CD, indicating the shift is actually 2 (the position of B). The candidate must recognise that the shift value is derived from the first pair’s internal structure, then apply that derived shift to the third term. This self‑referential layer forces the solver to extract a parameter from the first analogy before executing the second.

Mechanically, recursive analogies can be modelled as a two‑stage function: f(x) = g(h(x), x) where h extracts a parameter from the first term and g applies it. In letter‑based items, h often counts letters, measures alphabetical distance, or detects symmetry (palindrome, mirror). In number‑based items, h may compute digital root, prime‑index, or the number of distinct prime factors. The second stage g then uses that parameter as an operator (addition, multiplication, rotation). Because the parameter is not given explicitly, the solver must test candidate parameters against the given second term until a unique fit emerges.

Common recursive patterns include: (1) “first‑letter shift” where the shift equals the alphabetical index of the first letter of the first term; (2) “length‑based scaling” where a numeric multiplier equals the number of letters in the first word; (3) “difference‑mirror” where the difference between the two terms of the first pair is reversed and applied to the third term. Each pattern can be nested: a recursive analogy may contain a second recursion, demanding three‑stage extraction. The key to speed is to list all observable attributes of the first pair (length, first/last letter, vowel count, digit sum) and treat each as a potential parameter. Once a parameter reproduces the second term exactly, it is locked in for the third term.

Examiners often disguise recursion by embedding it inside a longer chain. For instance, a three‑pair sequence 12 : 30 :: 18 : 42 :: 24 : ? may look like a simple linear function, but the step from 12 to 30 could be “multiply by 2 then add 6”, where the multiplier 2 comes from the number of distinct prime factors of 12 (2 and 3 → count 2) and the addend 6 comes from the sum of those primes. The same parameter extraction applied to 18 (prime factors 2,3 → count 2, sum 5) yields 18×2+5=41, not 42, so the rule is rejected. The correct recursion might be “multiply by the number of letters in the English spelling of the first number”. Twelve has six letters → 12×2+6=30; eighteen has eight letters → 18×2+8=44 (not 42). The solver must iterate through plausible attribute functions until the chain holds for all given pairs.

Practice with recursive items builds the habit of treating the first analogy as a data source rather than a static example. Write a quick attribute table for the first term: alphabetical indices, letter‑type counts, numeric properties. Then test each attribute as a parameter against the second term. The attribute that yields an exact match becomes the rule for the remainder of the question.

Matrix‑Based Analogy Grids with Row‑Column Operations

Matrix analogies present a two‑dimensional array of letters, numbers, or symbols where the relationship operates along rows, columns, or both simultaneously. The classic format shows a 3×3 grid with one missing cell, but analogy questions often give two complete grids and ask for the missing element in a third grid that follows the same row‑column rule. The rule can be decomposed into independent row operations and column operations that combine (e.g., row‑wise addition modulo 26, column‑wise multiplication). Because the grid supplies multiple data points, the solver can verify a hypothesised rule against several cells before committing.

Row operations frequently involve sequential shifts: each cell in a row is derived from its left neighbour by a fixed Caesar shift, a reverse‑alphabet mirror, or a numeric increment. Column operations may apply a different shift, a scaling factor, or a logical operation such as XOR on binary representations of alphabet positions. When both dimensions are active, the value at row i, column j equals f(row‑rule(i), column‑rule(j)). A common composite is value = (row‑base + column‑offset) mod 26 for letters, or value = row‑base × column‑factor for numbers. The solver should first test pure row consistency (does each row follow a single pattern?) and pure column consistency. If both hold, the intersection cell is over‑determined and the rule is confirmed.

Advanced grids introduce wrap‑around at the edges (Z→A, 9→0) and non‑linear functions such as squaring the column index before adding. Some grids use a keyword to vary the column offset per column (Vigenère‑style). For instance, a 4×4 letter grid with keyword “MATH” (13,1,20,8) adds the keyword value to each column’s base letter. The first row might be the keyword itself, and subsequent rows apply a constant row shift. The missing cell is then solved by identifying the keyword from the first row, confirming the row shift from the second row, and projecting to the target coordinates.

Numeric grids often embed arithmetic progressions in rows and geometric progressions in columns, or vice‑versa. A 3×3 grid where the first row is 2, 4, 8 (×2) and the first column is 2, 6, 18 (×3) forces the cell (2,2) to be 4×3=12 or 6×2=12 — both agree, confirming the dual‑progression rule. If the two projections disagree, the grid follows a more complex function such as value = row‑base × column‑base + constant. The constant can be extracted from any known cell and then verified across the grid.

To solve matrix analogies efficiently, map the grid coordinates, write the observed values, and compute row‑differences and column‑differences. Look for constant differences, constant ratios, or alternating patterns. When a keyword is suspected, test the first row or column as the keyword by converting letters to numbers and checking whether subsequent rows/columns are derived by a uniform operation. The redundancy of the grid (usually 8 known cells for a 3×3) makes false hypotheses easy to discard.

Worked Example: Recursive Letter Shift
Given the analogy: CAT : FDW :: DOG : ? Options: (A) GRJ (B) GRK (C) GQJ (D) GQK
Step 1: Analyse the first pair CAT → FDW. Write alphabet positions: C=3, A=1, T=20 → F=6, D=4, W=23. Differences: +3, +3, +3. The shift is uniform +3. Step 2: Check whether the shift parameter is derived from CAT. CAT length = 3, first letter C=3, vowel count = 1. The uniform shift +3 matches the word length (3) and also the first‑letter index (3). Both are plausible. Step 3: Test each hypothesis on the third term DOG. DOG length = 3 → shift +3 gives GRJ (D+3=G, O+3=R, G+3=J). First‑letter index of D = 4 → shift +4 gives HSK (not an option). Therefore the governing parameter is word length. Step 4: Apply shift +3 to DOG → GRJ. Option (A) matches. Answer: (A) GRJ.
Exam‑Day Strategy
When a recursive analogy appears, immediately list three attributes of the first term: length, first‑letter index, and vowel/consonant count. Test each as the shift or multiplier against the second term. The attribute that reproduces the second term exactly is the rule; discard the others. This three‑candidate filter solves >80 % of recursive items within 15 seconds.

Semantic Field Shifts in Word Analogies

Semantic field analogies rely on relationships between words that belong to the same conceptual domain but differ in a systematic attribute such as intensity, size, temporal order, or hierarchical level. Unlike simple synonym/antonym pairs, these analogies require the solver to identify the underlying semantic dimension and then apply the same dimensional shift to the target word. For example, whisper : shout :: drizzle : ? operates on the intensity dimension of sound and precipitation respectively. The shift is from low intensity to high intensity within the same field. The answer “downpour” mirrors the high‑intensity member of the precipitation field.

The examiner constructs these items by selecting a semantic field (e.g., temperature, speed, social rank, cognitive effort) and two ordered levels within it. The first pair establishes the direction (low→high or high→low). The second pair supplies the base term from a different field but the same ordinal position (low). The candidate must retrieve the high‑intensity counterpart from that second field. Mastery demands familiarity with common taxonomic hierarchies: military ranks (private → general), academic degrees (bachelor → doctorate), geological time (epoch → era), musical dynamics (pianissimo → fortissimo), and so on.

Complex variants introduce a two‑step semantic shift: the first pair moves across two dimensions (e.g., intensity and valence). Consider glad : ecstatic :: angry : ?. Glad→ecstatic is positive valence, high intensity. Angry is negative valence, medium intensity. The target must be negative valence, high intensity → “furious” or “enraged”. The solver must decompose the first shift into its component dimensions and replicate both on the second base term. Another variant uses a “field‑crossing” shift: seed : oak :: larva : ?. The relation is developmental stage → mature form within the same biological kingdom (plant vs insect). The answer “butterfly” (or “beetle” depending on larva type) requires domain knowledge of life cycles.

To solve efficiently, first label the semantic dimension(s) of the given pair using a concise tag: “intensity↑”, “size↓”, “time→later”, “rank↑”. Then scan the options for a word that occupies the same tagged position in its own field. If multiple options fit, check for a secondary dimension (valence, animacy, concreteness) that distinguishes them. Building a mental catalogue of 30‑40 common fields with their ordered levels (e.g., coffee sizes: ristretto, espresso, lungo, americano; computer storage: bit, byte, kilobyte, megabyte, gigabyte) dramatically reduces retrieval time.

Examiners often embed a distractor that shares the correct field but the wrong ordinal level (e.g., “sprinkle” for drizzle→downpour). The trap is avoided by verifying that the option’s ordinal index matches the shift magnitude observed in the first pair. If the first pair jumps two levels (private → captain), the second must also jump two levels (apprentice → journeyman, not apprentice → master). Counting levels explicitly prevents the “adjacent‑level” error.

Algorithmic Generation of Number Series Analogies

Number series analogies present a pair of numbers where the second is produced by a deterministic algorithm applied to the first. The algorithm may be a composition of elementary arithmetic functions (addition, multiplication, exponentiation), digit‑level operations (reversal, sum of digits, product of digits), or number‑theoretic functions (prime counting, divisor sum, Euler’s totient). The analogy format x : y :: z : ? asks for the output of the same algorithm on a new input z. Because the algorithm is not given, the solver must reverse‑engineer it from the single example (x, y). This is an under‑determined problem; the test designer ensures uniqueness by restricting the algorithm to a known catalogue of “standard” functions used in reasoning exams.

The standard catalogue includes: (1) polynomial maps such as f(n)=n²+1, f(n)=2n+3; (2) digit‑manipulation like reverse digits, sum of squares of digits, concatenation of n and n+1; (3) prime‑based functions: next prime, nth prime, prime index of n, sum of prime factors; (4) divisor functions: number of divisors τ(n), sum of divisors σ(n); (5) base‑conversion tricks: interpret n in base b, convert to decimal; (6) recursive sequences: Fibonacci(n), triangular numbers Tₙ = n(n+1)/2. The exam expects candidates to recognise these patterns quickly.

Reverse‑engineering proceeds by computing a battery of candidate functions on x and checking which yields y. Because the function set is small, a mental checklist works: compute x², x³, 2ˣ, x!, sum of digits, product of digits, reverse(x), next prime after x, τ(x), σ(x). If one matches y, test it on z. If multiple match, the test designer has either made the item ambiguous (rare) or expects the simplest function (Occam’s razor). Simplicity is judged by the number of operations and the magnitude of constants. For instance, 12 : 144 :: 15 : ? could be square (12²=144) or multiply by 12 (12×12). Square is simpler (single operation, no extra constant) → answer 225.

Advanced items chain two functions: f(g(x)). Example: 5 : 13 :: 7 : ?. Candidates: 5→13 could be 2×5+3, or 5²−12, or Fibonacci(5)=5 then +8. The simplest linear function 2x+3 fits. Apply to 7 → 17. If 17 is an option, select it. If not, test the next simplest. Chains of three functions appear in higher‑difficulty sets; the solver must be prepared to compose functions mentally (e.g., “square, then sum digits, then add 1”). Writing the intermediate results for the given x helps verify each step.

Digit‑level algorithms often exploit the fact that the number of digits is small (2‑3 digits). For a 3‑digit x, operations like “reverse and add” or “cyclic shift of digits” are common. Example: 123 : 444 :: 456 : ?. 123 reversed 321, sum = 444. Apply to 456: reverse 654, sum = 1110. If options include 1110, that is the answer. When the result exceeds typical option range, the algorithm may include a modulo (e.g., take last three digits). Always check whether the given y is the raw result or a reduced form.

Practice with a curated list of 50 standard functions builds the pattern‑recognition speed needed to reverse‑engineer within 10‑15 seconds per item. Maintain a personal “function cheat sheet” grouped by type (polynomial, digit, prime, divisor) and rehearse applying each to random numbers.

Cross‑Domain Analogy Mapping Between Verbal and Quantitative Logic

Cross‑domain analogies require the solver to translate a relationship expressed in one symbolic system (verbal, numeric, or spatial) into another system while preserving the structural mapping. A typical item presents a verbal pair and a numeric pair that share an abstract relational skeleton, then asks for the missing element in a third pair that may be verbal, numeric, or mixed. For instance, hot : cold :: 100 : ? where the verbal relation is antonymy (opposite ends of a temperature scale) and the numeric relation is “subtract from a constant” (100 → 0). The constant 100 represents the boiling point of water in Celsius; the opposite end is 0. The answer 0 demonstrates mapping of the “opposite‑extreme” structure across domains.

The structural skeleton can be categorised into a small set of relational templates: (1) inversion (A ↔ opposite of A), (2) scaling (A → k·A), (3) translation (A → A + c), (4) composition (A → f(g(A))), (5) hierarchical step (A → parent/child in a taxonomy), (6) cyclic shift (A → next in a cycle). The solver must first identify the template from the fully given pair (verbal‑verbal or numeric‑numeric). Then apply the same template to the incomplete pair, respecting the target domain’s operations.

Verbal‑to‑numeric mapping often uses a quantifiable attribute of the words: length, alphabetical sum, vowel count, Scrabble score, or position in a known ordered list (months, planets, Mohs hardness). Example: January : 1 :: September : ?. The template is “month → ordinal number”. The answer 9. A more subtle item: cat : 3 :: dog : ? where the attribute is “number of distinct letters” (c,a,t = 3; d,o,g = 3) → answer 3. If the attribute were “alphabetical sum” (3+1+20=24; 4+15+7=26) the answer would differ. The test designer ensures only one attribute yields a plausible answer among the options.

Numeric‑to‑verbal mapping reverses the process: given a numeric relation, find a word pair that mirrors it. Example: 8 : 4 :: ? : ? with options like “square : root”, “cube : square”, “double : half”. The numeric relation is “divide by 2” (or “square root of square”). The verbal pair “double : half” expresses the same inverse scaling. The solver must recognise that “double” and “half” are multiplicative inverses, just as 8 and 4 are related by factor 2.

Mixed‑domain chains extend this to three or more terms: 2 : 4 :: apple : ? :: 5 : ?. The first numeric pair suggests “multiply by 2”. The verbal pair must follow the same template: a word that can be “doubled” in a meaningful sense — perhaps “apple : appleapple” (concatenation) or “apple : pear” (next fruit in a list). The third numeric pair then applies the same template to 5 → 10. The correct answer set will be the only one where all three pairs share a single abstract template. Solving such chains requires writing the template in symbolic form (e.g., f(x)=2x) and testing each domain’s candidate operations against it.

To develop cross‑domain fluency, practice translating a dozen common relational templates into at least three domains each (verbal, numeric, spatial). For “inversion”, verbal: hot/cold, numeric: x/‑x, spatial: left/right. For “cyclic shift”, verbal: Monday/Tuesday, numeric: mod 12 addition, spatial: compass points. This multi‑domain drilling makes the abstract skeleton visible regardless of surface symbols.

  • Assuming the first numeric operation that fits (e.g., square) is the only rule without testing simpler linear alternatives.
  • Ignoring wrap‑around in alphabetic shifts and treating Z→A as a break in the pattern.
  • Confusing word‑index (position inside the word) with alphabet‑index (A=1) in hybrid coding.
  • Applying a single‑step rule to a multi‑step chain and selecting an answer that matches only the first step.
  • Overlooking prime‑index or Fibonacci‑index relationships when numbers are small but not perfect squares or cubes.
  • Treating a reverse‑order letter pair as a simple Caesar shift instead of a mirror (sum = 27) pattern.
  • Failing to verify that the derived rule works for all given positions before committing to an option.

Consistent practice with these advanced patterns builds the pattern‑recognition speed required for high‑stakes reasoning sections.

Practice Questions

Coding-Decoding →