NNAT3 Level G common mistakes
NNAT3 Level G · Understand the traps and how to avoid them.
Treating the missing-patch options as independent pictures and picking the one that looks like the overall design, rather than checking that the patch's edges line up with the lines and colours that touch the hole.
The eye is drawn to the global pattern, so a distractor that repeats the motif but is rotated or colour-swapped a quarter-turn looks 'right' at a glance. NNAT3 patch items are designed so the global gestalt is satisfied by several options and only edge continuity separates them.
What to do instead
Ignore the whole figure for a moment and trace only the four borders of the hole. For each option, ask: does the line entering at the left edge leave at the same height? Does the colour band at the top continue at the same width? Eliminate any option that breaks a single border contact before judging the interior.
Choosing the patch option that is the correct shape but the wrong scale, so the pattern's stripes or checks change width across the join.
Candidates match shape and orientation but stop there, because scale feels like a minor cosmetic difference. In fact NNAT3 distractors often keep the shape and rotate or stretch it slightly, and the only tell is that the repeating unit no longer matches the surrounding spacing.
What to do instead
Count the repeating unit. If the visible area shows three stripes in a given span, the patch must show the same stripe width across the same span. Do a quick finger-measure of one repeat in the intact area and compare it with the option's repeat.
Solving the A-to-B pair by overall similarity ('both have a dot inside a shape') instead of stating the exact transformation in words before looking at the options.
Analogy items reward a verbal rule. Without one, candidates match on surface features such as colour, number of elements, or filled versus hollow, and pick a distractor that preserves those but changes the actual relationship (for example, moves the dot outside instead of inside).
What to do instead
Before viewing the answer choices, say the rule aloud in a single sentence: 'the small shape moves from the corner to the centre and flips colour.' Then apply that sentence to the third figure and only afterwards compare with the options. If two options fit, the rule was underspecified: find the extra attribute that also changed.
Assuming the transformation runs in the same direction in both halves of the analogy, when the item actually reverses it (A adds an element, C removes one).
The left-to-right reading habit makes 'A is to B' feel like a forward operation that must repeat. Some NNAT3 analogy items use a relation that is symmetric in appearance but asymmetric in effect, and mirror-image distractors exploit this.
What to do instead
Test the rule backwards as a sanity check: from B, can you get back to A by undoing it? If the rule is 'add a dot', the reverse is 'remove a dot'. Then check which direction the third figure requires before applying it, and prefer the option that satisfies the relation in the direction the item demands.
Predicting the next figure in a series by extrapolating the last visible change only, ignoring changes that occurred earlier and then stopped.
Series items often combine two rules with different periods, for example a shape that rotates 90 degrees each step while its fill alternates every second step. Looking only at the final transition hides the alternating rule and produces a confident wrong answer.
What to do instead
Tabulate the sequence: write each frame's key attributes (position, orientation, fill, count) in a column and read down each attribute separately. Any attribute that changes on every step follows one rule; any that flips every other step follows another. Extend each column independently, then combine.
Continuing a numeric or counting pattern by the visible increment without checking whether the increment itself is changing.
The most obvious arithmetic (add two each time) is often the distractor, while the true series uses a second-order rule such as adding 2, then 3, then 4, or doubling and subtracting one. Candidates under time pressure lock onto the first difference they compute.
What to do instead
Always compute two levels of differences before choosing. If first differences are constant, the rule is linear; if they change, compute the differences of the differences. Also test a multiplicative hypothesis by dividing consecutive terms. Only then look at the options.
Folding the paper mentally in the direction that feels natural rather than the direction stated, so the punched holes are mirrored to the wrong side.
Paper-folding items depend entirely on which half folds onto which. Without a physical reference, people default to a left-over-right fold regardless of the instruction, and the resulting hole pattern is a mirror image of the correct one, which is always offered as a distractor.
What to do instead
Before folding, mark the crease and label the moving flap. Track one hole at a time through each fold, and at every fold reflect the hole across the crease line, not across the paper's centre. Count the holes at the end: the number of holes must equal the number of punches times two to the power of the number of folds.
Rotating a 3-D object in the wrong plane, confusing a turn about the vertical axis with a tumble about the horizontal axis.
Most people mentally rotate objects about the vertical axis far more fluently than about the horizontal one, so a top view is misread as a front view. NNAT3 spatial items include options that are correct for the wrong axis, and they look plausible because the object's silhouette is unchanged.
What to do instead
Anchor the object with two named features, such as a marked face and a protruding tab, and state where each ends up after the stated rotation. If the item gives a view from above, decide first which edge is nearest the viewer; that single decision eliminates most axis-confusion distractors.
Spending equal time on every item, so the last items in a timed section are reached with seconds to spare and answered by guessing.
NNAT3 is speeded and items are not ordered strictly by difficulty, so a hard item early can consume the budget for three easy ones later. Candidates who treat every item as equally costly routinely leave several unattempted.
What to do instead
Adopt a 30-second rule: if no answer is in sight after about half a minute, mark the best partial guess, flag the item, and move on. Return only if time remains. Practise with a visible timer so the pace becomes automatic rather than a conscious calculation.
Second-guessing a first correct answer after re-reading the options, and switching to a distractor that only differs by a detail the rule does not mention.
Under time pressure, re-checking without a criterion turns into a search for a reason to doubt. Distractors are built to look more 'complete' or more symmetrical than the correct answer, and anxiety favours the more elaborate option.
What to do instead
Re-check only against the stated rule, never against a feeling. Write the rule, verify the chosen option satisfies it, and confirm the rejected option violates a named part of it. If both survive, the rule is incomplete: identify the additional attribute that changed and re-evaluate once, then commit.