OLSAT Level A common mistakes
OLSAT Level A · Understand the traps and how to avoid them.
On verbal analogy items, the child picks the answer pair that is merely associated with the prompt pair (e.g., 'dog is to cat' answered with 'bone is to mouse') rather than the pair that repeats the same relationship (pet is to pet, or animal is to animal).
Level A testers are early readers who reason from real-world stories and habits. A familiar pairing like dog/bone feels 'true,' so the child stops checking whether the relationship in the stem is the same one being applied to the answer.
What to do instead
Teach a two-step routine the child says aloud: first name the stem relationship in a single sentence ('a dog and a cat are both pets'), then test each option by plugging it into that same sentence. Practice with five-item sets where two options are associated-but-wrong, so the child learns to reject 'sounds related' in favor of 'same rule.'
The child answers a 'which one does not belong' item by choosing the option that is personally disliked, unfamiliar, or scary (e.g., rejecting 'spider' from a group of insects because it is creepy, or rejecting 'carrot' from a group of vegetables because it is disliked), instead of the item that breaks the shared category.
Young children classify by emotional salience and personal experience before they classify by category membership. The odd item is often the one that draws attention, so attention and logic point in opposite directions.
What to do instead
Have the child state the category of the majority first ('these three are all ___'), then check each option against that label. Use sorting games with household objects where the 'odd one out' is neutral (three spoons and a sock) so the child practices the label-first habit without emotional interference.
On sentence-completion or 'which sentence makes sense' items, the child selects an option after reading only the first few words or looking at the picture, without finishing the sentence, so a sentence that starts plausibly but ends absurdly gets chosen.
Level A readers are still decoding, so reading is effortful and they conserve effort by sampling. Pictures also pull attention away from the final words that carry the deciding information.
What to do instead
Train a 'finger under the words, read to the period' rule, then a second pass asking 'does the whole sentence make sense?' Practice with deliberately mismatched items (e.g., 'The fish flew up into the ___') so the child experiences how the ending, not the beginning, decides the answer.
On verbal reasoning items that ask what must be true (e.g., 'All cats have whiskers; Fluffy is a cat. What do we know about Fluffy?'), the child answers from what is likely or what they know about the real world rather than from the statements given, and picks an option that is true in general but not forced by the premises.
Children this age reason concretely and draw on background knowledge automatically. The distinction between 'probably true' and 'must be true' is not yet a habit, and nothing in everyday conversation rewards it.
What to do instead
Use nonsense premises so background knowledge cannot help ('All bloms have three tails; Zib is a blom. How many tails does Zib have?'). This forces the child to use only the given information and makes the logic visible. Then move to real-world premises once the habit is set.
On pictorial classification items, the child groups pictures by color, size, or visual similarity (all the red things together, all the big things together) when the intended grouping is by function or category (things that fly, things you ride).
Pictorial items invite perceptual matching, and perceptual similarity is the most obvious feature available. The child has not been told that the rule can be about what objects do or are used for rather than how they look.
What to do instead
Before answering, have the child name each picture and its job ('this is a bus, it carries people'). Then ask which two share a job. Deliberately mix colors and sizes within categories so that visual matching produces wrong answers during practice and the child learns to look past appearance.
On pictorial sequencing items (what happens first, next, last), the child orders the pictures by where they sit on the page or by which picture looks most interesting, rather than by cause and effect, and so places the result before the action.
Left-to-right reading order is a strong habit, and many practice materials present sequences in order, so the child assumes position encodes sequence. Also, the most vivid picture (the finished cake, the broken vase) attracts the first choice.
What to do instead
Have the child narrate the story aloud with time words ('first... then... at the end...') before touching the answer. Practice with scrambled sequences where page position conflicts with logical order, and ask the child to justify with 'because' after each placement.
On figural analogy items (e.g., big circle to small circle as big square to ?), the child copies the change in only one attribute, matching the size change but ignoring the shape change, or matches the shape but drops the size change.
Young children center on one salient dimension at a time (classic centration). Tracking two simultaneous transformations exceeds what they do by default, especially when one change is more visually striking.
What to do instead
Teach a checklist said aloud: 'What changed? What stayed the same?' Have the child point to and name each attribute (shape, size, color, position, fill) in the stem pair before looking at options, then eliminate options that fail either test. Start with one-attribute changes and build to two.
On figural classification and matrix items, the child selects an answer that completes a pattern visually but violates the row or column rule, because they match the nearest neighbor instead of checking the whole row and the whole column.
Matrix items require holding two rules at once (across and down). The child latches onto the most recent or most adjacent figure and copies it, a local-matching strategy that works often enough on simple items to become a habit.
What to do instead
Cover the answer choices and have the child describe the row rule and the column rule separately before uncovering options. Use a finger to trace across the row and then down the column. Build matrices from 2x2 with one rule to 3x3 with two rules so the two-rule check becomes automatic.
The child rushes through the last items or guesses randomly when time pressure builds, leaving items blank or marking answers without reading all options, especially near the end of a timed section.
Level A candidates have limited stamina and weak time monitoring. Once they sense they are behind, anxiety replaces strategy, and young children often do not know that an unanswered item scores the same as a wrong one while a considered guess may score.
What to do instead
Practice short timed sets (five to eight items) so pacing is learned in small doses, and teach a simple rule: answer every item, and if stuck, pick the best remaining option and move on. Rehearse a 'two-pass' approach: do the easy items first, then return to hard ones.
The child misreads the question stem, answering 'which one is different' as 'which ones are the same,' or 'which comes next' as 'which came first,' and then solves the wrong problem correctly.
Direction words (different, same, next, first, not, except) are abstract and visually similar to young readers, and the child's momentum carries them into the options before the stem is fully processed.
What to do instead
Have the child underline or point to the key direction word in the stem and restate the task in their own words ('I need to find the one that is NOT like the others') before examining options. Drill with paired items that differ only in the direction word so the child sees how the answer flips.