Study guide
4.6 Shape Codes & Letter-Coded Figures
Kent Test 11+ Kent Selective Eligibility Test
Track your progress as you read
Finished sections get a check mark.
1 Chapter 1: English & Reading Comprehension
1.1 Reading Comprehension & Inference Strategies 1.2 Vocabulary, Synonyms, Antonyms & Context Clues 1.3 Grammar, Syntax & Sentence Structure 1.4 Punctuation, Spelling Rules & Proofreading 1.5 Sentence Completion: Choosing the Best Word
2 Chapter 2: Mathematics & Problem Solving
2.1 Numbers, Place Value, Decimals & Four Operations 2.2 Fractions, Percentages & Ratio 2.3 Sequences, Patterns & Algebra Foundations 2.4 Perimeter, Area & Compound Shapes 2.5 Data Handling, Probability & Word Problems 2.6 Factors, Multiples, Primes & Divisibility 2.7 Time, Money & Unit Conversion 2.8 Angles, Coordinates, 3D Shape & Volume
3 Chapter 3: Verbal Reasoning
3.1 Word-Based VR (Analogies, Synonyms & Odd One Out) 3.2 Word Building, Hidden Words & Compound Words 3.3 Codes, Ciphers & Letter Substitution 3.4 Logic, Letter Sequences & Statement Relationships 3.5 Double-Meaning Words & Two-Pair Matching 3.6 Number-Based Verbal Reasoning
4 Chapter 4: Non-Verbal & Spatial Reasoning
4.1 Shape Sequences, Series & Progressive Matrices 4.2 Shape Analogies, Similarities & Odd One Out 4.3 Spatial Reasoning: Rotation, Reflection & Symmetry 4.4 3D Spatial Reasoning, Nets & Cube Folding 4.5 Hidden & Embedded Target Shapes 4.6 Shape Codes & Letter-Coded Figures
5 Chapter 5: The Writing Exercise, Scoring & Exam Strategy
5.1 The Writing Exercise: Planning, Structure & Descriptive Technique 5.2 Kent Test Format, Age Standardisation & Qualifying Scores 5.3 Exam Strategy, Time Management & Answer Sheet Mechanics 5.4 Registration, Eligibility & Test-Day Logistics 5.5 Writing Prompts & the Headteacher Assessment Route
4.6 Shape Codes & Letter-Coded Figures
Key Takeaways Explain This Topic AI
The official familiarisation booklet notes that some non-verbal questions involve codes, where parts of a shape are matched with a letter or letters.
Each letter position in a code stands for one visual attribute, and letters in the same position always belong to the same attribute family.
The method is to build an attribute table from the coded examples, deduce what each letter means, and then read the target figure attribute by attribute.
Two-letter codes usually pair a shape attribute with a shading or line attribute; three-letter codes add a third such as size or the number of elements.
A code deduced from a single example is unsafe, because a letter's meaning is only confirmed when it recurs across figures that differ in other respects.
Last updated: August 2026
Shape Codes & Letter-Coded Figures
The official Kent Test familiarisation booklet, describing the non-verbal reasoning section, says that "some questions may also involve codes, where you have to match parts of a shape with a letter or letters." Code items are among the highest-value questions on the reasoning booklet: they look intimidating, so many children rush or guess, yet the underlying method is completely mechanical.
The Format
A set of figures is shown, each labelled with a two- or three-letter code. A new figure follows, and you must choose its code from the options.
The single governing principle is position :
Each position in the code stands for one attribute of the figure, and every letter appearing in that position describes a value of that same attribute.
So in a two-letter code, the first letter might describe the outer shape and the second the shading. A letter in the first position never describes shading, and a letter in the second position never describes the outer shape. That constraint is what makes the puzzle solvable.
The Attribute Table Method
Take a worked example. Four coded figures are given:
- Figure Description Code
- 1 Triangle, striped AX
- 2 Square, striped BX
- 3 Triangle, black AY
- 4 Circle, black CY
Build the table position by position.
Position 1 — compare figures that share this letter:
- A appears on figures 1 and 3. Both are triangles . → A = triangle
- B appears on figure 2, a square . → B = square
- C appears on figure 4, a circle . → C = circle
Position 2 — same process:
- X appears on figures 1 and 2. Both are striped . → X = striped
- Y appears on figures 3 and 4. Both are black . → Y = black
Now decode a target. A striped circle must be C (circle) then X (striped) → CX .
Notice how the deduction worked: A was confirmed by appearing on two figures that were different in shading but identical in shape . That is the logic to look for.
The Cross-Check That Prevents Errors
A letter meaning deduced from one figure is only a hypothesis. Figure 2 alone would allow B to mean "square" or "striped" — it cannot tell you which. The meaning is only pinned down when the letter recurs on a figure that differs in the other attributes, or when the position rule forces it.
So before answering:
Confirm each letter you intend to use appears in the position you are using it in. A letter borrowed from position 2 and used in position 1 is the single most common mistake.
Confirm each letter's meaning is supported by at least two figures , or by elimination within its position.
Decode the target in position order , writing the attribute names down before choosing letters.
Two-Letter and Three-Letter Codes
- Code length Typical attributes Example pairing
- Two letters Shape + shading Outer shape, then fill
- Two letters Shape + line style Outer shape, then solid/dashed border
- Three letters Shape + shading + size Outer shape, fill, then large/small
Three letters Shape + number of elements + orientation Outer shape, count of inner dots, then which way it points
With three-letter codes the attribute table has three columns rather than two, but nothing else changes. Work each column in isolation and never let a deduction in column 1 influence column 3.
- Attributes the papers actually use
- Outer shape and number of sides
- Inner shape, or the presence of an inner shape at all
- Shading: white, black, striped, dotted, cross-hatched
- Border line style: solid, dashed, dotted, double
- Size: large, medium, small
- Count: number of dots, lines, or repeated elements
- Orientation: which way a point, arrow or notch faces
- Position of an inner element: top, bottom, left, right, centre
- Worked Example With a Three-Letter Code
- Figure Description Code
- 1 Large white pentagon PLW
- 2 Small white pentagon PSW
- 3 Large black hexagon HLB
- 4 Small black pentagon PSB
Position 1: P appears on figures 1, 2 and 4, all pentagons ; H on figure 3, a hexagon .
Position 2: L on figures 1 and 3, both large ; S on figures 2 and 4, both small .
Position 3: W on figures 1 and 2, both white ; B on figures 3 and 4, both black .
A large white hexagon decodes as H + L + W = HLW .
Note that HLW never appeared in the examples. That is normal: the answer is usually a combination that has not been shown, which is exactly why guessing from visual similarity to a labelled figure fails.
Reverse Coding
Some items invert the task: you are given a code and must choose the figure it describes. The method is unchanged — build the same attribute table from the labelled examples, then read the code left to right and eliminate options attribute by attribute. Eliminating on the first letter usually removes three of the five options in one step, so always start there rather than looking for an overall match.
Speed Discipline
These items sit in the same 4-to-5-minute subtests as the rest of the non-verbal section. Build the attribute table on the booklet page — rough work in the booklet is permitted and is not marked — using two or three short columns of letters. Written deduction is faster and far more accurate than holding six letter-meanings in working memory, and it makes the cross-check almost free.