OLSAT Level B common mistakes
OLSAT Level B · Understand the traps and how to avoid them.
Treating a picture analogy (A is to B as C is to ?) as a story association, picking the option that 'makes sense together' rather than the option that repeats the exact same relationship.
Young test-takers reason by overall resemblance instead of isolating the relational link (part-to-whole, function, category, opposite). The first pair feels like a scene, so any plausible partner feels right.
What to do instead
Teach a two-step verbal routine: say the first pair's link out loud in a fixed sentence frame ('a ___ is a kind of ___', 'a ___ is used to ___', 'a ___ does not have ___'), then test each option against that same sentence and reject any option that needs a different sentence.
Answering a classification question (find the one that does not belong) by picking the item that is unusual, expensive, or personally disliked instead of the item that breaks the shared category.
Children substitute a salience judgement ('this one is weird') for a category judgement, and unusual-but-in-category items (a bat among birds) or familiar-but-out-of-category items (a spider among insects) both pull the wrong way.
What to do instead
Require a named rule for the majority before looking at options: 'All of these are ___.' Then check each option against that one rule, and for near-misses ask 'Is a bat a bird?' rather than 'Does a bat look like a bird?'
Losing track of word order in sentence completion and choosing an option that fits the meaning but not the grammar or the blank's position.
The candidate reads the sentence once, grabs the gist, and never re-reads with the chosen option inserted, so subject-verb agreement, singular/plural, and connector words (because, although, so) go unchecked.
What to do instead
Make 'insert and re-read the whole sentence' a mandatory last step, and pre-read the sentence to predict the kind of word needed (a naming word, an action word, or a joining word) before scanning options.
Stopping at the first option that seems to fit a verbal reasoning item and marking it without scanning the rest.
On timed sections, early confidence plus a desire to finish creates a first-plausible-answer bias; distractor options are deliberately written to be partly true.
What to do instead
Enforce a 'read all options, then choose' rule during practice, and require the candidate to cross out two options with a stated reason before marking. Track how often the first choice was actually correct to make the bias visible.
Guessing the pattern in a figure series from only the first two frames and then refusing to revise when later frames contradict the guess.
Two frames can support many rules (add one, rotate, alternate), and once a child has verbalised a rule they treat it as settled rather than as a hypothesis to be checked.
What to do instead
Teach the three-frame check: the rule must explain the change from frame 1 to 2 AND from 2 to 3, and must predict frame 4. Practise items with alternating and two-attribute rules so the candidate learns to look for a second changing feature.
Matching matrices and analogies by overall visual similarity (same shape, same colour, same size) instead of by the transformation between rows or columns.
Perceptual matching is the default strategy at this age; colour and size are more salient than position, count, or rotation, so a distractor that copies the look of the target gets chosen.
What to do instead
Force a verbal description of the rule before choosing: 'In each row the shape stays and the shading flips.' Then have the candidate point to the feature that changes and the feature that stays the same, and eliminate options that change the wrong feature.
Missing small count or quantity changes (number of dots, sides, or internal lines) in series and matrix items, choosing an option that is one step off.
Counting is effortful and children estimate quantity by sight; near-miss distractors differing by a single element are designed to catch estimation.
What to do instead
Build a habit of counting the variable feature aloud or with a finger tap on every frame before answering, and include practice sets where the only difference between the correct option and the best distractor is one element.
Ignoring rotation and orientation, treating a turned figure as a different figure, or accepting a mirrored version as the same.
Mental rotation is still developing, and mirrors are visually close to rotations, so left-right flips pass as 'the same shape turned'.
What to do instead
Practise with cut-out or drawn shapes physically rotated and flipped so the candidate feels the difference, and adopt the check 'turn it in your head - does it land exactly, or is it backwards?' before accepting an orientation option.
Rushing through the picture-classification and paper-folding-style items to bank time for reading-heavy items, then making avoidable errors on the visual items that were actually quick wins.
Candidates perceive nonverbal items as 'easy pictures' and verbal items as the 'real test', so they spend disproportionate time rereading verbal items and skim the visual ones.
What to do instead
Do a timed mixed practice set and show the candidate their own accuracy and time per domain; then set a per-item time budget for both domains and rehearse a pacing plan that gives visual items their full share.
Leaving blanks on difficult items instead of eliminating options and marking a best guess.
Fear of being wrong, plus not knowing that blank answers score the same as wrong answers, leads children to skip rather than guess after elimination.
What to do instead
Explain the scoring reality in plain terms, teach a fixed fallback routine (cross out clearly wrong options, then pick from what remains), and rehearse it in timed practice so the guess-and-move-on habit is automatic.