Non-Verbal Reasoning Techniques for Shape and Pattern Questions

“Non-verbal” does not mean you have to solve the question as a silent blur of shapes. One of the best ways to make a visual puzzle manageable is to turn what you see into a few compact rules.

Non-verbal reasoning techniques help you solve problems made from shapes, symbols, positions, rotations, sequences, and visual relationships instead of ordinary sentences. The challenge is usually not seeing the figures. It is deciding which properties matter and how those properties change.

A useful approach is to convert the display into a small set of trackable features: shape, number, orientation, position, size, shading, and relation. Then describe what happens to each one. “Rotate 90°,” “add one,” or “swap left and right” is much easier to test than a vague impression that two pictures somehow belong together.

This page focuses on technique. If you mainly want practice items, Cognitive Train's Non-Verbal Reasoning Questions page is the better starting point.

Technique 1: Break the Figure into Features

Do not treat a complex symbol as one indivisible object. Split it into attributes.

For a typical shape question, ask:

• What is the shape?
• How many elements are there?
• Where are they positioned?
• Which direction are they facing?
• Are they filled, outlined, striped, or differently colored?
• What relationship do the parts have to each other?

This decomposition helps because different attributes can follow different rules. A triangle may rotate clockwise while the number of internal dots increases. If you simply watch “the triangle,” those two changes can blur together.

Technique 2: Put the Visual Rule into Words

A surprisingly useful trick is to verbalize a visual transformation. Instead of repeatedly staring at two symbols, say: “the outer shape stays the same, the inner line rotates clockwise, and one dot is added.”

That does not make the task verbal in the ordinary sense. It gives the visual relationship a compact label that is easier to hold and compare.

Research on Raven-style reasoning supports the idea that people can rely on different representational codes. In an eye-tracking study, Srivastava, Jaarsveld, and Sangani distinguished more visuospatial and more verbal-analytical Raven problems and found different scanning strategies across the two types. Their results reinforce the broader point that non-verbal reasoning is not a single purely visual process.

So when a relation is hard to “see,” try naming it.

Technique 3: Compare Key Landmarks, Not Every Pixel

When one shape transforms into another, focus on distinctive landmarks: a sharp corner, a protruding arm, a notch, the longest edge, or the relative position of two parts.

This can be much faster than mentally comparing every contour point.

In experiments on shape transformation, Schmidt and Fleming asked observers to identify corresponding points between shapes before and after complex non-rigid transformations. Participants were strikingly consistent, and a model based on high-salience contour landmarks predicted their responses well. The study was not an aptitude test, but it provides a useful perceptual principle: salient shape features can anchor correspondence across transformation.

For reasoning questions, that means asking, “Where did the distinctive corner go?” can be more useful than asking, “Does this whole shape look similar?”

Technique 4: Separate Identity from Transformation

A common error is to treat a rotated object as a different object. Track what the shape is separately from what happened to it.

If the same asymmetric symbol appears turned 90 degrees, its identity has stayed constant while orientation changed. If the symbol is mirrored, that is a different transformation again.

Cognitive Train's Mental Rotation Test isolates orientation changes, while the Shape Assembly Test focuses more on how parts combine into larger forms. Both are useful neighboring skills when spatial transformation is the part of non-verbal reasoning that slows you down.

Technique 5: Look for Relations, Not Just Features

Some questions cannot be solved by tracking individual attributes. You have to notice a relationship: one figure is inside another, two shapes swap positions, the third cell combines the first two, or each row contains one of every symbol.

This is where relational reasoning becomes central. Carpenter, Just, and Shell's classic analysis of Raven's matrices found that stronger performance involved inducing abstract relationships and managing several problem-solving goals, rather than simply detecting isolated visual features.

The practical move is to write the relation mentally as an operation: A moves inside B, A + B = C, each row contains all three values. Once the operation is explicit, superficial changes matter less.

If this is the kind of reasoning you want to practice directly, Cognitive Train's Matrix Reasoning Test concentrates on relational patterns arranged across rows and columns.

Technique 6: Construct an Answer Before Eliminating

When possible, decide what the missing figure should look like before studying the answer choices closely.

This is known as constructive matching. It protects you from distractors that satisfy only part of the rule. Once you have a prediction, the choices become a verification step.

In a 2024 study covering participants from age 6 to 18, Gonthier and colleagues found that constructive matching increased strongly with age and was associated with better Raven performance. Strategy use mediated more than half of the age effect on reasoning performance in their sample.

If you cannot construct the whole answer, switch to elimination using whatever constraints you do know. “Must be mirrored” or “must contain three shapes” may be enough to remove most of the alternatives.

Technique 7: Track One Dimension at a Time

Hard visual questions often combine rules. Trying to see them all simultaneously overloads the problem.

Use a simple checklist:

• Shape: same or changing?
• Count: increasing, decreasing, or alternating?
• Orientation: rotating or reflecting?
• Position: moving through locations?
• Shading: cycling or distributed?

Finish one dimension before starting the next. Then combine the conclusions. If temporary rule tracking is consistently the difficult part, Cognitive Train's N-Back Test gives you a separate measure of working-memory updating.

Use the Simplest Rule That Explains Everything

A good non-verbal rule should survive more than one comparison. If “rotate clockwise” explains one pair but fails everywhere else, discard it. If a simpler rule explains every row, prefer it over a complicated story invented around one cell.

This is especially important when answer choices are tempting. A distractor often looks plausible because it matches one visible feature. Your rule needs to account for the full structure.

Practice the Techniques on Real Questions

Cognitive Train's Abstract Reasoning Test is the broadest live test for applying these techniques to visual-rule problems. The Visual Pattern Test is useful when the challenge is detecting shape-based regularities rather than following a verbal argument.

For another step-by-step guide, see How to Solve Abstract Reasoning Questions. The wider Pattern Recognition section connects these techniques with matrix reasoning, sequences, visual search, and memory.

You can also browse Cognitive Train's brain tests collection or explore the broader set of brain training and cognitive training tools from the homepage.