ABF Officer Recruit Cognify common mistakes
ABF Officer Recruit Cognify · Understand the traps and how to avoid them.
Rounding intermediate figures (e.g. treating 0.67 as 0.7 or 3.14 as 3) mid-calculation, then picking whichever answer option sits closest to the rounded result.
Under a per-question timer, rounding feels like free speed, and because all options are plausible-looking numbers, the nearest one creates a false sense of confirmation.
What to do instead
Estimate only to discard two options, then redo the arithmetic exactly on the surviving pair. Keep fractions or full decimals until the final line, and round only once, at the point of choosing.
Computing a correct number but answering the wrong quantity: giving percentage points when a percentage change was asked, or a per-day rate when the stem asked for a per-week rate.
The stem is read first, the calculation consumes all attention, and by the time the answer is produced the requested unit has dropped out of mind. Answer options are usually built to include the wrong-unit value.
What to do instead
Before calculating, write a one-line target on the scratchpad in the form 'answer = ___ units'. After calculating, read the target line before touching an option.
Selecting an option because it is true in general knowledge or sounds sensible, rather than because a specific sentence in the passage states or strictly implies it.
Passages are usually written on familiar topics, so prior knowledge fills gaps silently and feels indistinguishable from comprehension.
What to do instead
For every option, including the one you like, point to the exact line that supports it. If you cannot cite one, mark it as unsupported regardless of whether it is factually true.
Treating quantifying words as interchangeable, so that a passage saying 'some managers' is used to support an option saying 'most managers' or 'managers generally'.
The gist of a sentence survives in memory while the exact quantifier fades, and the wrong option reads as a fair paraphrase of the gist.
What to do instead
Underline or note every quantifier (all, most, some, few, none, only) as you read, and test each option by matching its quantifier to the passage's before judging meaning.
Mentally rotating a shape and then choosing a mirror-image distractor, because the rotated figure and its reflection look equally familiar after several rotations.
Rotation without a fixed reference point loses handedness: once the shape is turned 90 degrees or more, the brain matches on overall outline and stops tracking which side a distinguishing feature sits on.
What to do instead
Pick one asymmetric landmark (a notch, a single dot, a longer tab) and state its position relative to the shape's centre before and after the rotation. Rule out any option where that landmark is on the opposite side.
Trying to rotate the whole object along all three axes at once, losing track halfway and re-starting the rotation two or three times per item.
Full 3D rotation exceeds what most people can hold in visual working memory, so the mental image degrades and the candidate restarts instead of narrowing.
What to do instead
Rotate along one named axis at a time and lock in an intermediate orientation. Also eliminate options that differ only by a single face or edge before attempting any rotation.
Scanning a search field in an ad-hoc zigzag, then re-checking areas already cleared or skipping a band entirely; the miss is usually in the same region as the last distractor fixated on.
Targets that appear after a near-miss distractor capture attention, which pulls the gaze off the systematic path, and there is no record of where the scan has already been.
What to do instead
Impose a fixed sweep order (left to right, top to bottom, in bands) and keep the cursor or a finger anchored at the current position. Never jump to a new region because something looked interesting.
Rehearsing an entire list or sequence repeatedly as one long string instead of grouping it, so the last two items overwrite the first two when a new item arrives.
Verbal rehearsal holds only a few seconds of content, and repeating the whole string costs more time per cycle exactly when the sequence is longest.
What to do instead
Chunk items into groups of three or four with a rhythmic grouping, and rehearse only the current group plus a one-word tag for each earlier group. If the task allows, drop each item as soon as it has been used.
Forming a rule from the first two figures in a series and applying it immediately, without testing it against the third and fourth figures.
Early items are the most salient and almost any rule fits two data points, so the first hypothesis feels confirmed when it is only unrefuted.
What to do instead
State the candidate rule, then verify it against the third and fourth items before answering. Check one attribute at a time in a fixed order: count, position, rotation, shading, then element type.
Changing answers during a final review sweep with no new evidence, or conversely spending so long eliminating options that the last items are answered blind.
Time pressure at both ends of the section triggers two opposite habits: over-deliberation on early items and panic revision on later ones, and neither is driven by the question content.
What to do instead
Use a per-item time cap and a two-pass rule: eliminate clearly wrong options, commit, and move on. Only revisit a flagged item if you can name a specific fact you misread the first time.