NNAT3 common mistakes
NNAT3 · Understand the traps and how to avoid them.
On Pattern Completion items, choosing the answer that matches the overall look of the large design without checking whether the missing piece's edges line up with the lines, shading, or shapes touching the hole.
The eye is drawn to the global pattern, and the answer options are deliberately designed so that one option looks 'right' at a glance but is subtly misaligned at the boundary. Candidates compare the whole picture instead of the seam.
What to do instead
Cover the answer choices, then trace each line or shaded region that touches the edge of the missing area and predict where it must continue. Only after predicting the edge behaviour, look at the options and eliminate any that break a single traced line.
On Pattern Completion, assuming the hidden piece is a mirror image of the surrounding design rather than a continuation of it, so the chosen piece is flipped.
Symmetry is common in these designs, so candidates apply a symmetry rule that does not actually hold across the gap. The option that is a mirror of the correct piece is usually present as a distractor.
What to do instead
Test one asymmetric feature (a diagonal line, an off-centre dot, a shade that appears on only one side). If the design is not symmetric about the gap, the piece must continue the direction of the surrounding lines, not reflect them. Practise by identifying whether each design is symmetric before choosing.
On Reasoning by Analogy, matching on a single salient feature (for example, both figures are triangles, or both are dark) and declaring the pair complete without checking the transformation.
The most visually striking shared attribute captures attention first, and the analogy format encourages a fast 'these two go together' judgement. Distractors are built to share exactly that salient feature.
What to do instead
State the relationship out loud in a full sentence of the form 'A changes to B by ___, ___ and ___', naming at least two attributes (shape, count, size, orientation, shading, position). Then apply the same sentence to C and reject any option that satisfies only one of the named attributes.
On Reasoning by Analogy, applying the A-to-B change in the wrong direction or inverting it, so the chosen answer is the inverse transformation of the one required for C.
Candidates often reason from B back to A instead of from A forward to B, especially when B is the more memorable figure. The inverse rule produces a plausible-looking but wrong option.
What to do instead
Physically point from A to B and describe the change as a forward action ('the shape loses one side and rotates 90 degrees clockwise'). Then point from C and perform that same forward action. Check by confirming the answer could not be obtained by running the rule backwards.
On Reasoning by Analogy, ignoring the position or arrangement of elements inside the figure and tracking only the outer shape, so an answer with correct shapes in the wrong arrangement is selected.
Outer contour is processed faster than internal layout, and working memory load makes it easy to drop the internal detail after the shape has been matched.
What to do instead
Adopt a fixed scan order for every figure: outer shape, then number of internal elements, then their positions (top/bottom/left/right/centre), then shading, then orientation. Compare options against the full checklist rather than stopping at the first match.
On Serial Reasoning, extrapolating the sequence by eye from the first two or three frames instead of counting the step size, so a rule that is actually 'add two each time' is treated as 'add one'.
The sequence is presented left to right and the brain completes the visual rhythm quickly. Small step sizes look identical at a glance, and the distractor for the wrong step size is always offered.
What to do instead
Write the numeric value of the changing attribute under each frame (for example, 2, 4, 6, 8). Compute the difference between consecutive frames explicitly and confirm it is constant or follows a stated progression before selecting the next frame.
On Serial Reasoning, assuming every attribute changes across the series when in fact some attributes are constant, so the candidate over-transforms and picks a frame where a fixed feature has moved.
Once a candidate finds one changing attribute, they generalise and start looking for change everywhere. Constant features are easy to overlook because attention is on the motion.
What to do instead
Before solving, list every attribute and mark each as 'changes' or 'stays the same' by comparing frame 1 with frame 2. Only the attributes marked 'changes' may differ in the answer; any option that alters a constant attribute is eliminated immediately.
On Serial Reasoning with two interleaved rules (alternating or paired patterns), treating the series as a single progression and therefore predicting the wrong next frame.
Alternating sequences hide two rules inside one row, and the candidate who checks only adjacent frames sees an inconsistent pattern and guesses. The correct answer often requires comparing frames one apart.
What to do instead
Check for alternation by comparing frame 1 with frame 3 and frame 2 with frame 4. If two separate consistent rules emerge, apply the rule for the odd or even positions to determine which rule governs the missing frame, and verify the predicted frame against its same-parity neighbours.
On Spatial Visualization, mentally rotating the whole figure as a rigid object when the item actually requires folding, reflecting, or combining parts, so the answer is a rotated copy of the original rather than the required transformation.
Rotation is the most practised mental operation, so it becomes the default strategy. Reflection in particular is easy to mistake for rotation because both preserve shape and size.
What to do instead
Name the required operation before looking at options: rotation, reflection, folding, or part-assembly. If it is a reflection, track one asymmetric marker (a letter, a dot, a notch) and confirm it ends up on the opposite side, since rotation would keep it on the same side. Use your own hand or the paper edge as a physical mirror check where permitted.
On Spatial Visualization, losing track of a fold line or overlap so the candidate counts a shape twice, or places it on the wrong side of the crease, when predicting the unfolded result.
Folding problems require holding several positions in working memory simultaneously, and the crease line is often not visually emphasised. Errors accumulate silently because the final picture still looks plausible.
What to do instead
Work backwards from the folded state one fold at a time, drawing or tracing the position of a single marked feature at each unfold. Confirm the feature crosses to the mirror position relative to the crease, and check the final count of features against the number of layers created by the folds.