Study guide
5.1 Number Matrices & Grid Patterns
Edutest Selective Entry & Scholarship Exam Exam
Track your progress as you read
Finished sections get a check mark.
1 Chapter 1: Exam Structure, Results & Pacing Strategy
1.1 Exam Structure, Format & Test Battery 1.2 Results, Year-Level Norms & How Schools Select 1.3 Speed Tactics, Time Budgeting & Zero-Penalty Guessing
2 Chapter 2: Verbal Reasoning: Analogies, Classification & Vocabulary
2.1 Word Analogies & Semantic Relationships 2.2 Verbal Classification & Odd-One-Out 2.3 Advanced Vocabulary, Synonyms & Antonyms
3 Chapter 3: Verbal Reasoning: Codes, Sequences & Logical Deduction
3.1 Letter Codes & Cryptographic Shifts 3.2 Word Chains, Anagrams & Hidden Words 3.3 Verbal Deductive Logic & Syllogisms
4 Chapter 4: Numerical Reasoning: Sequences & Number Patterns
4.1 Arithmetic, Geometric & Two-Tier Differences 4.2 Alternating, Interleaved & Fibonacci Sequences 4.3 Square, Cube & Figurate Number Sequences
5 Chapter 5: Numerical Reasoning: Grids, Diagrams & Math Logic
5.1 Number Matrices & Grid Patterns 5.2 Geometric Shape Puzzles & Relational Diagrams 5.3 Cryptarithms & Quantitative Word Logic
6 Chapter 6: Reading Comprehension: Fiction & Literary Analysis
6.1 Narrative Structure, Plot Devices & Conflict 6.2 Figurative Language, Imagery & Tone 6.3 Characterization, Inferencing & Subtext
7 Chapter 7: Reading Comprehension: Non-Fiction, Poetry, Argument & Sentence Skills
7.1 Informational Text Structures & Main Idea 7.2 Poetry Analysis: Form, Rhyme, Rhythm & Meaning 7.3 Persuasive Techniques, Rhetoric & Author Bias 7.4 Sentence Correction, Sentence Completion & Punctuation
8 Chapter 8: Mathematics: Arithmetic, Fractions, Decimals & Percentages
8.1 Mental Arithmetic, Order of Operations & Number Properties 8.2 Fractions, Decimals & Mixed Operations 8.3 Percentages, Percentage Change & Financial Problems
9 Chapter 9: Mathematics: Rates, Ratios, Probability & Data
9.1 Ratios, Rates & Proportional Reasoning 9.2 Speed, Distance, Time & Combined Work 9.3 Probability, Counting Principles & Permutations 9.4 Data Interpretation, Dual-Axis Charts & Statistics
10 Chapter 10: Mathematics: Measurement, Geometry & Algebra
10.1 Perimeter, Area & Composite 2D Figures 10.2 Volume, Surface Area & 3D Spatial Visualization 10.3 Angles, Geometric Theorems, Polygons & Transformations 10.4 Algebraic Expressions, Linear Equations & Word Problems
11 Chapter 11: Written Expression: Narrative & Creative Writing
11.1 Narrative Architecture & Deconstructing Prompts 11.2 Sensory Immersion & Show-Don't-Tell Techniques 11.3 Mechanics, Punctuation Variety & Under-the-Gun Polish
12 Chapter 12: Written Expression: Persuasive & Expository Writing
12.1 Persuasive Prompt Deconstruction & Thesis Formation 12.2 High-Scoring Body Paragraphs: The PEEL/TEEL Framework 12.3 Expository Clarity, Rhetorical Flourish & Clincher Closures
5.1 Number Matrices & Grid Patterns
Key Takeaways Explain This Topic AI
Number matrices (commonly 3 × 3 3 \times 3 3 × 3 grids) evaluate multidirectional pattern recognition, requiring candidates to distinguish row rules, column rules and whole-grid invariants quickly.
Directional triage begins by analysing numerical gradients: horizontal expansion indicates row-wise operations ( C 1 ∘ C 2 = C 3 C_1 \circ C_2 = C_3 C 1 ∘ C 2 = C 3 ), vertical expansion indicates column-wise operations ( R 1 ∘ R 2 = R 3 R_1 \circ R_2 = R_3 R 1 ∘ R 2 = R 3 ), and uniform dispersion suggests global constant sums.
The core arithmetic archetypes comprise constant sum/product invariants, linear combinations ( C 3 = a ⋅ C 1 + b ⋅ C 2 C_3 = a \cdot C_1 + b \cdot C_2 C 3 = a ⋅ C 1 + b ⋅ C 2 ), multiplicative/fractional products, power/difference relations, and digit-sum transformations.
The 20-Second Grid Triage Protocol prioritises hypotheses systematically: (1) constant row/column sums, (2) element 1 & 2 operations to produce element 3, (3) digit-level sums, and (4) diagonal or whole-grid invariants.
Never execute a calculated guess based on a single row; verifying the proposed rule across a second independent row or column is mandatory to eliminate deceptive distractor traps.
Last updated: October 2026
5.1 Number Matrices & Grid Patterns
At a Glance: Numerical matrices test your multidirectional arithmetic reasoning under intense time pressure. Unlike linear number sequences that progress along a single timeline, matrices conceal rules along rows, down columns, across diagonals, or as whole-grid invariants. Mastering the 20-Second Grid Triage Protocol enables you to determine the operational direction instantly, identify the arithmetic archetype, and eliminate distractors in 25–30 seconds.
Number Matrices in Numerical Reasoning
Edutest names matrices as one of the four skills in its Numerical Reasoning test. The test lasts 30 minutes, and Edutest states that actual tests contain between 40 and 60 questions, so the average pace is about 30 to 45 seconds per question . Edutest does not publish how many items are matrices. Matrix puzzles are designed to test fluid intelligence and mental arithmetic agility without relying on complex curriculum-specific mathematics.
In standard test batteries, you will encounter three principal grid formats:
3 × 3 3 \times 3 3 × 3 Standard Matrices: Nine cells arranged in three rows and three columns, with a question mark ( ? ? ? ) typically occupying the bottom-right corner ( R 3 C 3 R_3C_3 R 3 C 3 ) or the central cell ( R 2 C 2 R_2C_2 R 2 C 2 ).
4 × 4 4 \times 4 4 × 4 Extended Matrices: Sixteen cells featuring longer progressions, often operating under two-step progressive rules or compound row-and-column sums.
Segmented / Irregular Grids: Multi-cell rectangular grids (such as 2 × 3 2 \times 3 2 × 3 or 3 × 4 3 \times 4 3 × 4 ) where elements operate in linked pairs or triplets.
When confronted with a grid under timed exam conditions, unpractised students waste 40 to 60 seconds randomly performing trial-and-error calculations across various cells. High-scoring candidates, by contrast, follow an invariant diagnostic triage protocol that systematically eliminates dead ends in seconds.
Operational Directionality: Decoupling Rows, Columns, and Global Invariants
The fundamental challenge of any number matrix is determining its operational vector : which cells combine to produce which result? There are three structural directions: Row-Major Rules Column-Major Rules Global Invariants C 1 ∘ C 2 = C 3 R 1 ∘ R 2 = R 3 ∑ Row = ∑ Col = K (Horizontal progression) (Vertical progression) (Whole-matrix property) \begin{array}{ccc}
- \text{\textbf{Row-Major Rules}} & \text{\textbf{Column-Major Rules}} & \text{\textbf{Global Invariants}} \\
- C_1 \circ C_2 = C_3 & R_1 \circ R_2 = R_3 & \sum \text{Row} = \sum \text{Col} = K \\
- \text{(Horizontal progression)} & \text{(Vertical progression)} & \text{(Whole-matrix property)}
\end{array} Row-Major Rules C 1 ∘ C 2 = C 3 (Horizontal progression) Column-Major Rules R 1 ∘ R 2 = R 3 (Vertical progression) Global Invariants ∑ Row = ∑ Col = K (Whole-matrix property)
The Variance Gradient Rule
To determine the operational vector within five seconds of viewing a grid, scan the variance gradient (the rate of numerical growth across rows versus columns):
Horizontal Growth Gradient: If numbers within each row start small on the left and grow substantially larger toward the right (e.g. Row 1: 4 , 7 , 33 4, 7, 33 4 , 7 , 33 ; Row 2: 6 , 3 , 27 6, 3, 27 6 , 3 , 27 ), the operational direction is almost certainly row-major ( C 1 ∘ C 2 = C 3 C_1 \circ C_2 = C_3 C 1 ∘ C 2 = C 3 ).
Vertical Growth Gradient: If numbers in the top row are modest but expand significantly in the bottom row (e.g. Column 1: 8 , 12 , 48 8, 12, 48 8 , 12 , 48 ; Column 2: 6 , 14 , 42 6, 14, 42 6 , 14 , 42 ), the operational direction is column-major ( R 1 ∘ R 2 = R 3 R_1 \circ R_2 = R_3 R 1 ∘ R 2 = R 3 ).
Uniform Magnitude (No Gradient): If all numbers across the grid share a similar magnitude (e.g. all values range between 12 and 19), look immediately for constant sums across rows, columns, or the whole matrix.
Reading Order Invariants (Snake / Book Sequences)
Occasionally, a matrix does not operate as independent rows or columns. Instead, the entire grid represents a single linear sequence folded into a rectangular container. These fall into two scanning paths:
Book Reading Order: Moving left-to-right across Row 1, dropping to Row 2, then Row 3 ( R 1 C 1 → R 1 C 2 → R 1 C 3 → R 2 C 1 → … R_1C_1 \to R_1C_2 \to R_1C_3 \to R_2C_1 \to \dots R 1 C 1 → R 1 C 2 → R 1 C 3 → R 2 C 1 → … ).
Serpentine (Boustrophedon) Order: Moving left-to-right across Row 1, right-to-left across Row 2, and left-to-right across Row 3 ( R 1 C 1 → R 1 C 2 → R 1 C 3 → R 2 C 3 → R 2 C 2 → R 2 C 1 → … R_1C_1 \to R_1C_2 \to R_1C_3 \to R_2C_3 \to R_2C_2 \to R_2C_1 \to \dots R 1 C 1 → R 1 C 2 → R 1 C 3 → R 2 C 3 → R 2 C 2 → R 2 C 1 → … ).
Note
If numbers steadily increase or decrease across the entire grid without clear row or column formulas, immediately test the reading order sequence for an arithmetic difference or geometric ratio.
Matrix Pattern Taxonomy: The Five Core Archetypes
The table below outlines five common mathematical relationships used in number matrices:
Pattern Archetype Algebraic Signature Diagnostic Signature Quick Test Protocol
1. Constant Sum / Product ∑ j = 1 3 C j = K \sum_{j=1}^3 C_j = K ∑ j = 1 3 C j = K or ∏ j = 1 3 C j = K \prod_{j=1}^3 C_j = K ∏ j = 1 3 C j = K Uniform numbers, no extreme outliers Add Row 1; add Row 2. Do sums match?
2. Linear Combinations C 3 = a ⋅ C 1 ± b ⋅ C 2 C_3 = a \cdot C_1 \pm b \cdot C_2 C 3 = a ⋅ C 1 ± b ⋅ C 2 Moderate growth in element 3 Test ( C 1 + C 2 ) × k (C_1 + C_2) \times k ( C 1 + C 2 ) × k or a C 1 + C 2 a C_1 + C_2 a C 1 + C 2
3. Product & Quotient C 3 = C 1 × C 2 k C_3 = \frac{C_1 \times C_2}{k} C 3 = k C 1 × C 2 Element 3 is divisible by factors of C 1 , C 2 C_1, C_2 C 1 , C 2 Multiply C 1 × C 2 C_1 \times C_2 C 1 × C 2 ; compare to C 3 C_3 C 3
4. Power & Difference C 3 = C 1 2 − C 2 C_3 = C_1^2 - C_2 C 3 = C 1 2 − C 2 or C 3 = C 1 2 + C 2 2 C_3 = C_1^2 + C_2^2 C 3 = C 1 2 + C 2 2 Element 3 is large; elements 1 & 2 are small Square C 1 C_1 C 1 ; compare difference to C 2 C_2 C 2
5. Digit-Sum Invariants sum_digits ( C 1 ) = sum_digits ( C 2 ) \text{sum\_digits}(C_1) = \text{sum\_digits}(C_2) sum_digits ( C 1 ) = sum_digits ( C 2 ) Seemingly random two-digit numbers Sum digits of each cell in Row 1
The 20-Second Grid Triage Protocol
When you reach a matrix question, run this structured four-phase triage protocol. Do not deviate or spend more than five seconds on an unfruitful phase. +-------------------------------------------------------------------------+
- | THE 20-SECOND GRID TRIAGE PROTOCOL |
- +-------------------------------------------------------------------------+
- | Phase 1: Rapid Sum Check (0 - 4 seconds) |
- | * Mentally sum Row 1: R1 = c1 + c2 + c3 |
- | * Mentally sum Row 2: R2 = c1 + c2 + c3 |
- | * If R1 == R2, constant sum found! Apply to Row 3. |
- | * If R1 != R2, check Col 1 sum vs Col 2 sum. |
- +-------------------------------------------------------------------------+
- |
- v (Sums differ)
- +-------------------------------------------------------------------------+
- | Phase 2: Extremal Combination Check (5 - 11 seconds) |
- | * Examine the row/column with the largest terminal number. |
- | * Test direct addition: C1 + C2 = C3? |
- | * Test scaled addition: (C1 + C2) * k = C3? |
- | * Test linear weighting: a * C1 + C2 = C3 or C1 * C2 - k = C3? |
- | * Verify hypothesis against Row 2 immediately! |
- +-------------------------------------------------------------------------+
- |
- v (No elementary linear relation)
- +-------------------------------------------------------------------------+
- | Phase 3: Product & Power Relations (12 - 17 seconds) |
- | * Test product-to-quotient: (C1 * C2) / k = C3? |
- | * Test squares/powers: C1^2 - C2 = C3 or C1^2 + C2 = C3? |
- | * Test mean/average: (C1 + C2) / 2 = C3? |
- +-------------------------------------------------------------------------+
- |
- v (Numbers appear chaotic/irregular)
- +-------------------------------------------------------------------------+
- | Phase 4: Digit-Sum / Positional Fallback (18 - 22 seconds) |
- | * Sum the individual digits of each cell (e.g. 48 -> 4 + 8 = 12). |
- | * Check reading order sequence (Row 1 -> Row 2 -> Row 3). |
- | * Check prime, square, or parity (odd/even) properties. |
- +-------------------------------------------------------------------------+
Tip
The Two-Row Verification Rule: Never assume you have cracked the matrix after testing only Row 1. Multiple algebraic formulas can fit any single pair of numbers (e.g. 4 4 4 and 7 7 7 can yield 28 28 28 via 4 × 7 4 \times 7 4 × 7 , or 23 23 23 via 4 × 4 + 7 4 \times 4 + 7 4 × 4 + 7 , or 33 33 33 via ( 4 + 7 ) × 3 (4+7) \times 3 ( 4 + 7 ) × 3 ). Always verify your candidate formula on Row 2 before applying it to Row 3.
- Step-by-Step Worked Grid Examples
- Worked Example 1: Third-Element Linear Combination
Problem: Determine the missing number ( ? ? ? ) in the 3 × 3 3 \times 3 3 × 3 grid below: 5 8 23 4 7 19 6 9 ? \begin{array}{|c|c|c|}
- \hline
- 5 & 8 & 23 \\
- \hline
- 4 & 7 & 19 \\
- \hline
- 6 & 9 & ? \\
- \hline
- \end{array} 5 4 6 8 7 9 23 19 ?
Step 1: Execute Phase 1 (Sum Check).
Row 1 sum: 5 + 8 + 23 = 36 5 + 8 + 23 = 36 5 + 8 + 23 = 36 .
Row 2 sum: 4 + 7 + 19 = 30 4 + 7 + 19 = 30 4 + 7 + 19 = 30 .
The row sums differ, so there is no constant row sum. Column sums also differ: Column 1 is 5 + 4 + 6 = 15 5 + 4 + 6 = 15 5 + 4 + 6 = 15 and Column 2 is 8 + 7 + 9 = 24 8 + 7 + 9 = 24 8 + 7 + 9 = 24 .
Step 2: Execute Phase 2 (Combination Check).
In Row 1, look for a way to combine 5 5 5 and 8 8 8 into 23 23 23 . Simple addition gives 13 13 13 , which is too small, so test a multiplier: 3 × 5 + 8 = 15 + 8 = 23 3 \times 5 + 8 = 15 + 8 = 23 3 × 5 + 8 = 15 + 8 = 23 .
Candidate rule: C 3 = 3 ⋅ C 1 + C 2 C_3 = 3 \cdot C_1 + C_2 C 3 = 3 ⋅ C 1 + C 2 .
Step 3: Verify on Row 2 before using the rule. 3 ⋅ ( 4 ) + 7 = 12 + 7 = 19 3 \cdot (4) + 7 = 12 + 7 = 19 3 ⋅ ( 4 ) + 7 = 12 + 7 = 19
The rule holds on a second, independent row.
Step 4: Apply to Row 3. C 3 = 3 ⋅ ( 6 ) + 9 = 18 + 9 = 27 C_3 = 3 \cdot (6) + 9 = 18 + 9 = \mathbf{27} C 3 = 3 ⋅ ( 6 ) + 9 = 18 + 9 = 27
Why the second row matters: the pair ( 5 , 8 ) (5, 8) ( 5 , 8 ) alone could also fit other rules, such as 2 × 8 + 7 = 23 2 \times 8 + 7 = 23 2 × 8 + 7 = 23 . That alternative predicts 2 × 7 + 7 = 21 2 \times 7 + 7 = 21 2 × 7 + 7 = 21 for Row 2, not 19 19 19 , so the verification step rejects it.
Worked Example 2: Column-Major Product & Quotient
Problem: Find the value of ? ? ? in the following numerical grid: 8 6 9 12 14 16 48 42 ? \begin{array}{|c|c|c|}
- \hline
- 8 & 6 & 9 \\
- \hline
- 12 & 14 & 16 \\
- \hline
- 48 & 42 & ? \\
- \hline
- \end{array} 8 12 48 6 14 42 9 16 ?
Step 1: Determine operational direction.
Scan the variance gradient:
Row 1 has single-digit values ( 8 , 6 , 9 8, 6, 9 8 , 6 , 9 ).
Row 2 has values in the teens ( 12 , 14 , 16 12, 14, 16 12 , 14 , 16 ).
Row 3 has values in the forties ( 48 , 42 , ? 48, 42, ? 48 , 42 , ? ).
The gradient runs vertically down columns. The operational direction is column-major : R 1 R_1 R 1 and R 2 R_2 R 2 combine to generate R 3 R_3 R 3 .
Step 2: Formulate the functional relationship in Column 1.
We must combine 8 8 8 and 12 12 12 to yield 48 48 48 .
Direct multiplication: 8 × 12 = 96 8 \times 12 = 96 8 × 12 = 96 .
Notice that 96 / 2 = 48 96 / 2 = 48 96/2 = 48 .
Candidate rule: R 3 = R 1 × R 2 2 R_3 = \frac{R_1 \times R_2}{2} R 3 = 2 R 1 × R 2 .
Step 3: Validate against Column 2.
Test with Column 2 values ( R 1 = 6 , R 2 = 14 R_1 = 6, R_2 = 14 R 1 = 6 , R 2 = 14 ): R 3 = 6 × 14 2 = 84 2 = 42 R_3 = \frac{6 \times 14}{2} = \frac{84}{2} = 42 R 3 = 2 6 × 14 = 2 84 = 42
The rule is verified across two independent columns.
Step 4: Execute on Column 3.
Apply to R 1 = 9 , R 2 = 16 R_1 = 9, R_2 = 16 R 1 = 9 , R 2 = 16 : R 3 = 9 × 16 2 = 9 × 8 = 72 R_3 = \frac{9 \times 16}{2} = 9 \times 8 = \mathbf{72} R 3 = 2 9 × 16 = 9 × 8 = 72
The missing value is 72 72 72 .
Worked Example 3: Digit-Sum Invariant Matrix
Problem: What number completes the pattern in the matrix below? 39 48 57 28 64 19 75 84 ? \begin{array}{|c|c|c|}
- \hline
- 39 & 48 & 57 \\
- \hline
- 28 & 64 & 19 \\
- \hline
- 75 & 84 & ? \\
- \hline
- \end{array} 39 28 75 48 64 84 57 19 ?
Step 1: Test standard arithmetic operations.
Row 1 differences: 48 − 39 = + 9 48 - 39 = +9 48 − 39 = + 9 ; 57 − 48 = + 9 57 - 48 = +9 57 − 48 = + 9 . (Row 1 is an arithmetic progression of + 9 +9 + 9 ).
Row 2 differences: 64 − 28 = + 36 64 - 28 = +36 64 − 28 = + 36 ; 19 − 64 = − 45 19 - 64 = -45 19 − 64 = − 45 . (Row 2 completely breaks the progression).
Row sums: Row 1 sum = 144 = 144 = 144 . Row 2 sum = 111 = 111 = 111 . Row sums do not match.
Column sums: Column 1 sum = 142 = 142 = 142 . Column 2 sum = 196 = 196 = 196 . Columns do not match.
Step 2: Apply Phase 4 (Digit-Level Invariants).
Calculate the sum of the digits for each number in Row 1:
- 39 ⟹ 3 + 9 = 12 39 \implies 3 + 9 = 12 39 ⟹ 3 + 9 = 12
- 48 ⟹ 4 + 8 = 12 48 \implies 4 + 8 = 12 48 ⟹ 4 + 8 = 12
- 57 ⟹ 5 + 7 = 12 57 \implies 5 + 7 = 12 57 ⟹ 5 + 7 = 12
All numbers in Row 1 possess an internal digit sum of 12 .
Calculate the digit sums for Row 2:
- 28 ⟹ 2 + 8 = 10 28 \implies 2 + 8 = 10 28 ⟹ 2 + 8 = 10
- 64 ⟹ 6 + 4 = 10 64 \implies 6 + 4 = 10 64 ⟹ 6 + 4 = 10
- 19 ⟹ 1 + 9 = 10 19 \implies 1 + 9 = 10 19 ⟹ 1 + 9 = 10
All numbers in Row 2 possess an internal digit sum of 10 .
Calculate the digit sums for Row 3:
- 75 ⟹ 7 + 5 = 12 75 \implies 7 + 5 = 12 75 ⟹ 7 + 5 = 12
- 84 ⟹ 8 + 4 = 12 84 \implies 8 + 4 = 12 84 ⟹ 8 + 4 = 12
The missing number must also yield a digit sum of 12 .
Evaluating candidate options with digits summing to 12 12 12 (e.g. 93 93 93 , where 9 + 3 = 12 9 + 3 = 12 9 + 3 = 12 ) cracks the problem instantly without any complex multi-variable algebra.
Distractor Traps & Speed Verification Tactics
Selective entrance test developers construct multiple-choice distractors around predictable student calculation errors. Recognizing these traps prevents costly mistakes:
The Single-Operation Trap: Calculating an operation on Row 1 (e.g. 4 + 7 = 11 4 + 7 = 11 4 + 7 = 11 ) and immediately selecting 8 + 5 = 13 8 + 5 = 13 8 + 5 = 13 on Row 3, ignoring the fact that Row 1's actual terminal entry was 33 33 33 .
The Inverted Column Trap: Applying a row rule horizontally when the actual relationship runs vertically. Always double-check whether the first column fits the rule before executing.
Parity Screening Shortcut: Before computing exact figures, check the odd/even parity. In linear combination rules like C 3 = 2 C 1 + 4 C 2 C_3 = 2 C_1 + 4 C_2 C 3 = 2 C 1 + 4 C 2 , C 3 C_3 C 3 must be even regardless of inputs. If three of the four multiple-choice options are odd, you can identify the correct response in two seconds flat.