CBEST common mistakes
CBEST · Understand the traps and how to avoid them.
Choosing an answer choice because it is true in general or matches personal experience, rather than because the passage supports it and it answers the specific question stem.
Prior knowledge and opinion intrude easily, and under time pressure the first plausible-sounding option feels safe, so the candidate stops evaluating once something 'sounds right'.
What to do instead
Before looking at the options, restate the stem in your own words and predict the kind of answer needed. Then for each option you are tempted by, physically point to the sentence or paragraph in the passage that supports it; if you cannot locate one, eliminate the option regardless of how true it seems.
Missing stems that ask for the exception or the negative - EXCEPT, NOT, least likely, does NOT support - and selecting the one choice that actually fits the passage.
The eye skips small function words during fast reading, and the brain processes the topic of the question while discarding its logical direction.
What to do instead
Circle or underline the negative word the moment you read the stem. Then label each option T (supported by the passage) or F (not supported) in the margin and select the single option that carries the opposite label from the other three.
Over-reading inference items: selecting the most dramatic, absolute, or far-reaching choice, or importing an assumption the passage never makes.
Inference feels like guesswork, and confidently worded options (always, never, all, must, proves) sound like the product of strong reasoning when they are usually traps.
What to do instead
Require a two-step chain for any inference: the passage states X, therefore Y follows, and Y must be a small step. Eliminate any choice containing absolute language unless the passage itself is absolute, and eliminate any choice that requires information from outside the passage.
Mistranslating word problems into equations - writing '5 less than 3 times a number' as 5 - 3x, setting up percent increase as a simple subtraction, or reversing the ratio in a rate problem.
Candidates match keywords to operations rather than reading for meaning, and percent and rate problems involve two different possible bases or denominators that look interchangeable.
What to do instead
Assign a variable and translate the sentence phrase by phrase, in order, before computing. Test the expression with a friendly number (let the number be 10) to see whether it produces a sensible result. For percent change always write (new - old) divided by old, and for rates write the units as a fraction so the correct arrangement is visible.
Losing accuracy inside multi-step arithmetic: misaligned decimals, mixing units (minutes with hours, centimeters with meters), and forgetting to convert before calculating.
Each step feels easy in isolation, so the candidate stops writing work down and carries intermediate values in their head, where units and place values silently drift.
What to do instead
Write every step on its own line with the unit attached to each number, convert everything to a single unit before the first calculation, and estimate the rough answer first so a result that is off by a factor of ten or sixty is caught immediately.
Conflating paired concepts whose names or formulas blur together: mean versus median, area versus perimeter, combinations versus permutations, independent versus mutually exclusive events.
These pairs are studied close together and stored as vocabulary rather than as distinct procedures, so recall produces the wrong formula under time pressure.
What to do instead
Build a two-column card for each pair with a one-line definition in your own words plus a tiny worked example. Add a dimensional check: if the answer should describe a surface it carries square units; if it should describe a distance it does not.
Leaving items blank, refusing to estimate, and grinding on one hard question until several later, easier questions go unanswered.
Candidates assume a wrong answer is penalized and believe that a hard item early in the section signals that the rest will be hard too.
What to do instead
There is no penalty for an incorrect response, so every item must have a marked answer before time expires. Cap time per item, use estimation or plug the answer choices back into the problem to back-solve, flag anything unfinished, and return to it only after the whole section has been attempted.
Writing the wrong genre: turning an analytical or opinion prompt into a personal narrative, or turning a personal-experience prompt into an abstract policy argument with no story.
The narrative mode is the most practiced and most comfortable, so under a 30-minute clock the writer defaults to storytelling even when the task asks for a position and support.
What to do instead
Before planning, underline the task verb and the audience and say aloud what kind of essay is required. For the analytical prompt, write a one-sentence thesis in the first paragraph and make each body paragraph one reason plus one example that supports that thesis.
Filling the essay with vague generalities and abstract claims ('communication is important in every workplace') instead of specific, concrete detail.
Writers try to sound formal and authoritative, and abstraction is easier to generate than remembered detail, so the essay inflates without gaining content.
What to do instead
Spend the first two minutes listing two or three specific incidents or examples with names, times, places, and numbers. Build each body paragraph around one of those concrete items; a two-sentence specific example scores better than a paragraph of abstractions.
Losing conventions under time pressure: run-ons and comma splices, shifting verb tense mid-narrative, no paragraph breaks, and no time left to proofread.
The essay is handwritten with no spell-checker and written against a clock, so sentence boundaries are the first thing to collapse once the writer is racing the final paragraph.
What to do instead
Reserve the last three to four minutes and proofread only for sentence boundaries: check that every sentence has one subject-verb core and one end mark, that joined clauses are joined by a conjunction or a period, and that each new idea starts a new paragraph.