11+ non-verbal reasoning is the subject that surprises children most on first encounter. There are no words to read, no sums to calculate — just shapes, patterns, and visual puzzles that test pure logical thinking. For some children it clicks immediately; for others it is initially baffling. Either way, NVR is one of the most trainable parts of the 11+, and focused NVR 11+ practice on shapes and patterns builds speed fast. This guide covers every core question type, exam technique, and a preparation plan from first encounter to exam day.
What is non-verbal reasoning?
Non-verbal reasoning (NVR) tests a child's ability to identify patterns, relationships, and rules in visual information — shapes, diagrams, grids, and sequences — without using language or numbers. It was originally designed to measure raw reasoning ability independently of vocabulary or learned knowledge, on the premise that a child who has not had the benefit of a rich reading environment could still demonstrate strong logical thinking.
In practice, NVR is learnable — children who practise systematically improve substantially, regardless of their starting point. The question types are finite and well-documented, the rules governing each type are logical, and the skills required can be built through focused, regular practice.
NVR appears in the 11+ for many grammar schools, particularly those using GL Assessment papers — including all Buckinghamshire grammar schools, which test all four subjects (Buckinghamshire 11+ — NVR is tested). Warwickshire 11+ — also tests NVR, like Bucks and unlike Kent. In CEM-format exams such as the Birmingham 11+, NVR is blended with mathematics in Paper 2. Trafford is a notable exception among CEM areas — NVR is not tested there. If your child is preparing exclusively for Trafford grammar schools — NVR not tested, NVR preparation can be deprioritised. In all other areas where NVR is tested, the underlying skills are the same.
The key visual properties NVR tests
Every NVR question is built around a finite set of visual properties. Children who learn to notice and track these properties systematically solve questions far faster than those who approach each question from scratch.
Shape type
Triangle, square, circle, pentagon, hexagon — what category is the shape?
Always name the shape first before looking at other properties.
Size
Small, medium, large — does size change across the sequence?
Compare shapes relative to each other, not in absolute terms.
Shading / fill
Empty, hatched, grey, or solid black — how is the shape filled?
Shading often cycles — white, grey, black, white again.
Rotation
How many degrees has the shape turned — 45°, 90°, 180°?
Use clock positions to describe rotation — “pointing to 3 o’clock.”
Reflection
Is the shape mirrored horizontally, vertically, or diagonally?
Physically flip the page or trace to check — don’t guess on reflections.
Number of elements
How many dots, sides, lines, or shapes are present?
Count carefully — adding or subtracting one element per step is a common rule.
Position
Where is the element within its frame — which corner, edge, or centre?
Map positions using compass points: NW, N, NE, W, centre, E, SW, S, SE.
Line style & direction
Solid, dashed, horizontal, vertical, diagonal — what type of lines are present?
Lines inside shapes often change independently of the shape’s outline.
Teaching your child to scan for these properties in order — every time, on every question — is the single most powerful NVR technique there is. It transforms an overwhelming visual puzzle into a methodical checklist. Most questions change only two or three properties at a time; a child who checks systematically will find the rule within thirty seconds. The SCSNRPL method below turns this list into a fixed routine your child can say aloud in the exam.
Every NVR question type explained
There are nine core NVR question types. Every question your child will face in the 11+ fits into one of these categories.
Sequences
Very common — appears in almost every paper
A row of shapes changes according to a rule. Find the next shape in the sequence, or identify the missing shape within it.
Check each visual property in turn. The rule usually involves one or two changing properties — rotation, shading, or number of elements.
Matrices
Very common — a core question type in all formats
A 2×2 or 3×3 grid of shapes follows rules both across rows and down columns. One cell is missing — find the shape that completes the grid.
Check the rule across rows first, then down columns. Both rules must work simultaneously for the correct answer.
- Rule across: empty → black
- Rule down: empty → grey
- Answer: black
Odd one out
Common — tests ability to spot the grouping rule
Five shapes are shown. Four share a property or rule; one does not belong. Identify the odd one out.
Look for the grouping that covers exactly four shapes. Try each visual property in turn — the answer is always clear once you find the right one.
Analogies
Common — tests whether your child can apply a rule consistently
Shape A is to shape B as shape C is to which option? The same transformation (size, shading, rotation, etc.) must apply to the second pair.
State the rule in words first — “small empty becomes large empty” — then apply it to shape C. Check every answer option against that rule.
Same change: size up, shading unchanged
Reflections
Common — mirror-line questions appear in most GL papers
Which option is the reflection of the original shape across a vertical, horizontal, or diagonal mirror line? Reflection flips handedness; rotation does not.
Mark one asymmetric feature (a dot, notch, or arrow tip) and trace where it lands after the flip. Eliminate any option that is a rotation instead of a mirror image.
Left and right swapped — not rotated
Rotations
Common — often paired with reflection as a distractor
Which shape shows the figure turned 45°, 90°, 135°, or 180° without being flipped? The shape keeps the same handedness; only its orientation changes.
Use clock positions for an asymmetric mark — “dot at 12 o’clock” becomes “dot at 3 o’clock” after 90° clockwise.
90° clockwise — same handedness
Nets and 3D shapes
Regular — cube nets are the most frequent
Which flat net folds to make the 3D shape shown, or which cube can be made from the net? Counting faces eliminates wrong answers before visualising the fold.
Count faces first — a cube needs exactly six. Only then try to picture which edges meet when the net folds.
6 faces = cube net
Shape codes
Common in GL papers — logic-table style decoding
Shapes are paired with symbols or letters according to a rule. Work out what each property means, then apply the code to a new shape.
Build a small table on rough paper: list each shape feature (triangle, shaded, dot) and its symbol. Test the rule on every given pair before answering.
The 10 best NVR tips for the 11+
These are the techniques that consistently separate well-prepared children from those who are guessing. Each one is learnable — none requires innate spatial ability.
Use the SCSNRPL checklist on every question
Before guessing, scan for Shape, Count, Size, shading (colouring), Number of elements, Rotation, Position, and Line style — in that order. Most rules involve only two of these. The checklist turns visual chaos into a methodical search.
Works on all question typesUse clock positions for rotation
Instead of trying to count degrees, describe the orientation of an asymmetric element using clock positions. “The arrow points to 3 o’clock.” After a 90° clockwise rotation, it points to 6 o’clock. This is faster and less error-prone than visualising abstract degree values.
Rotations · Sequences · AnalogiesNever confuse reflection with rotation
Exam setters deliberately include rotations among the answer options for reflection questions, and vice versa. The distinction: a reflected shape is a mirror image — left and right (or top and bottom) are swapped. A rotated shape is the same “handedness” spun around. Mark one asymmetric detail and trace where it ends up.
Reflections · RotationsFor matrices, check rows and columns independently
In a 3×3 matrix, a rule runs across every row and a (possibly different) rule runs down every column. Check the row rule first, then verify the column rule. The correct answer must satisfy both simultaneously. Answers that satisfy only one of the two rules are the most common traps.
MatricesCount faces before visualising nets
For net questions, count the number of faces in the net and the number of faces on the 3D shape before attempting to visualise the fold. A cube needs exactly 6 faces; a triangular prism needs 5. Any net with the wrong face count is immediately eliminated — no spatial reasoning required.
Nets and 3D shapesAssume one or two rules, not ten
Setters usually change only a couple of properties per step — rotation plus shading, or size plus number of dots. Once you have found two properties that explain the pattern, stop searching. Children who hunt for a fifth hidden rule often talk themselves out of the correct answer.
Sequences · Odd one outUse rough paper freely
Sketch rotations, mark mirror lines, draw clock positions, and note which SCSNRPL letters you have checked. Practise this in mocks so it feels normal in the exam — children who are shy about rough paper lose marks they could easily have secured on shape codes and reflection questions.
Works on all question typesEliminate wrong options before guessing
In five-option questions, crossing out two implausible answers turns a one-in-five guess into a one-in-three chance — and often better if you have spotted the setter’s trap. Wrong answers in NVR are rarely random; they usually break one obvious property such as face count, rotation direction, or shading.
Multiple choiceNever leave a question blank
Most 11+ NVR papers do not penalise wrong answers. A blank scores zero; an educated guess after eliminating two options has a real chance of a mark. Over 80 questions, always answering can be worth several marks — potentially the difference between qualifying and not.
Exam techniquePractise at exam pace and review by property
GL NVR is often 80 questions in 40 minutes — 30 seconds per question. Untimed success rarely transfers without timed training from around six months before the exam. After each paper, ask which letter of SCSNRPL was missed, not just “I got it wrong” — that targets the real gap for next time.
Timed practice · Error reviewThe SCSNRPL method: your child's secret weapon
The most powerful NVR technique is having a fixed checklist to apply to every question, every time. Experienced children do not stare at a shape hoping inspiration strikes — they scan through a mental checklist until the rule reveals itself.
The checklist covers the visual properties shown earlier in this guide. A useful mnemonic to remember them is SCSNRPL — "Some Children See Naturally, Reason Properly Later."
- S — Shape type. What is the basic shape — circle, triangle, square, pentagon, arrow?
- C — Colour and shading. Is it empty, grey, hatched, or solid black? Does shading change across the sequence?
- S — Size. Is the shape getting larger, smaller, or staying the same?
- N — Number of elements. How many shapes, dots, sides, or lines are there? Is the count changing?
- R — Rotation. Which direction is the shape pointing? Has it turned 45°, 90°, 135°, or 180°?
- P — Position. Where in the frame is the shape — which corner, edge, or region?
- L — Line style. Are the lines solid or dashed? Horizontal, vertical, or diagonal? How many are there?
Working through this checklist takes around 20–30 seconds on a new question. With practice it becomes automatic and takes far less. Teach your child to say the checklist aloud during practice — verbalising the process makes it stick.
How NVR is marked and scored
Most NVR questions in the 11+ are multiple choice with five options. There is no penalty for wrong answers in most papers — so children should always answer every question, even if they are guessing. An educated guess after eliminating two or three options gives a one-in-two or one-in-three chance of being correct, which is far better than a blank.
NVR scores are standardised alongside age, which means younger children in the year group are compared against norms for their age rather than against older peers. This partly offsets the disadvantage of being among the youngest in the year.
In GL papers, NVR typically runs to 80 questions in 40 minutes — which works out at exactly 30 seconds per question. This is tight. Children who have not practised under timed conditions consistently find the pace of the real exam significantly harder than untimed practice suggested. Timed practice from around six months before the exam is essential.
Preparation plan: from first encounter to exam day
Phase 1
Months 1–2
Introduce all nine question types
Work through each type with a structured workbook, one or two types per session. No time pressure — focus entirely on understanding the logic. Introduce the SCSNRPL checklist and practise saying it aloud. Identify immediately which types your child finds hardest.
Phase 2
Months 3–5
Build speed and target weak spots
Introduce gentle time pressure — aim for 45 seconds per question. Identify the two or three types where your child is slowest or least accurate and give them dedicated sessions. Continue spatial games and puzzles alongside formal practice. Begin mixing question types within sessions.
Phase 3
Months 6–9
Timed mixed practice papers
Move to full timed practice papers combining all question types. Review every wrong answer after each session — identify whether errors come from misapplying the rule, from time pressure, or from a specific question type that still needs work. Track progress across papers.
Phase 4
Final 6 weeks
Full mock exams under exam conditions
Sit full NVR mock papers at the correct speed — 30 seconds per question — in exam conditions. At least one full mock per week. In the final two weeks, ease back slightly. Arrive at exam day rested, confident, and familiar with the format rather than burned out from overpreparation.
Two to three NVR sessions per week of 30–40 minutes each is more effective than daily long sessions. Rest days allow spatial reasoning skills to consolidate — this is not unique to NVR; it applies to any skill that involves building new mental habits.
Pair formal practice with spatial play (below) from Phase 1 onward — it accelerates rotation, nets, and matrix skills without feeling like extra homework.
Spatial games that build NVR skills — without feeling like revision
One of the most underused NVR preparation strategies is play. The spatial and logical thinking skills that NVR tests develop through any activity that requires mentally manipulating shapes in space — and there are many that children genuinely enjoy.
Tangrams are particularly well matched to NVR. The challenge of fitting seven geometric pieces into a silhouette develops exactly the mental rotation and shape-fitting skills that nets, sequences, and hidden figure questions demand.
Minecraft and similar building games require constant 3D spatial reasoning — planning structures, visualising how layers stack, and rotating designs before placing them. Children who play these games regularly often develop stronger 3D visualisation ability than those who only do paper exercises.
Origami is one of the most direct analogues to net questions. Folding paper into 3D shapes and then unfolding it back develops an intuitive sense of how 2D surfaces become 3D objects — exactly the mental process net questions test.
Jigsaw puzzles at increasing difficulty levels develop the pattern recognition and shape-fitting skills underlying sequences and matrix questions.
Rubik's cubes and similar rotation puzzles build familiarity with 3D rotation — directly relevant to rotation and reflection questions.
None of these feel like exam preparation to a ten-year-old. But the cumulative benefit of an hour a week spent on spatial play alongside formal practice is measurable in NVR performance.
Common mistakes that cost marks — and how to fix them
- Scanning too quickly. The most expensive mistake in NVR is not spotting a changing property because the scan was too rushed. Train your child to check at least four or five properties before committing to an answer, even if the "obvious" answer seems clear within five seconds.
- Confusing rotation with reflection. Already covered in Tip 3, but worth repeating because it costs marks so consistently. Exam setters always include the rotation as a distractor in reflection questions. Pick one asymmetric feature and track it.
- Not checking both row and column rules in matrices. Many children find the row rule and answer from that alone, missing a constraint from the column rule that eliminates their chosen answer.
- Guessing on nets without counting faces. The simple act of counting faces eliminates two or three wrong answers in net questions without any spatial reasoning — yet most children do not do it.
- Forgetting that no answer penalty means always guess. Blank answers score zero. A guess after eliminating two options has a one-in-three chance of scoring a mark. Over a paper of 80 questions, this difference in approach can be worth several marks — potentially moving a child from below to above the qualifying threshold.
Recommended resources for NVR preparation
- Bond 11+ Non-Verbal Reasoning is the standard resource. The age 9–10 and 10–11 workbooks cover all question types clearly and progress from easier to harder questions within each type. Complete the 9–10 book first, regardless of your child's current year.
- CGP 11+ Non-Verbal Reasoning is well-presented and works well as a second resource once Bond is complete, providing additional variety in presentation and question style.
- Bond Assessment Papers (NVR) are the best source of timed mixed practice for Phase 3. Work up from the 9–10 age band through to 10–11 as confidence builds.
- Atom Learning covers NVR with adaptive difficulty and useful progress tracking. The platform's ability to identify which question types your child finds hardest and serve more of those specifically makes it time-efficient.
- Spatial reasoning puzzle books — available from most educational bookshops — provide valuable supplementary practice in a format that feels less formal. They are particularly useful for children who are resistant to sitting down for formal practice sessions.
Frequently asked questions about NVR
My child has no spatial awareness at all. Can they still improve?
Yes — this is one of the most well-evidenced findings in 11+ preparation. Spatial reasoning is a skill, not a fixed trait. Children who start with very weak NVR performance consistently make significant gains through focused, regular practice. The question types are finite and learnable; the SCSNRPL method gives even spatially unconfident children a reliable procedure to follow.
How is NVR different from non-verbal reasoning in the CEM exam?
In CEM papers, NVR is blended into a combined reasoning and mathematics paper rather than tested separately. The question types are a subset of the nine described above — typically sequences, matrices, and analogies feature most prominently. Because the CEM format varies year to year, broad preparation across all types is more reliable than trying to predict which will appear.
Is NVR harder than verbal reasoning?
For most children, NVR is easier to improve in quickly but harder to master at the highest level. The initial improvement curve is steep because the question types are unfamiliar rather than genuinely difficult. At the top of the ability range, NVR questions involving complex multiple-property rules or unusual rotation angles can be very challenging even for well-prepared children.
Should my child use rough paper during the NVR exam?
Absolutely — and this is worth practising explicitly. Drawing the SCSNRPL checklist, marking asymmetric features to track rotation, sketching out a decoding table for shape codes, and counting faces in net questions are all legitimately useful on rough paper. Children who are shy about using rough paper lose marks they could easily have secured.
How many NVR questions typically appear in the 11+?
For GL papers, NVR is usually a standalone paper of 80 questions in 40 minutes. For CEM papers, NVR questions are embedded within a broader reasoning paper — the exact number varies. For independent school exams, check the specific school's past papers for format details.