SIFT common mistakes
SIFT · Understand the traps and how to avoid them.
While working a Simple Drawings grid, the candidate loses track of which row or cell they last checked and either re-checks answered cells or silently skips a whole row, leaving several items blank at the end of the section.
The items are visually near-identical and the page is dense; the eye drifts diagonally instead of returning to a fixed left margin, so the internal 'cursor' resets to the wrong place after each mark.
What to do instead
Adopt a strict left-to-right, top-to-bottom sweep and physically anchor with the pencil tip or a finger resting on the left edge of the current row; touch each symbol once as you compare it, and at the end of every row do a two-second backward glance to confirm a mark exists in every cell.
On Hidden Figures, the candidate stares at the whole complex drawing looking for the complete target shape as a unit, spending 30-45 seconds on one item and then guessing.
Whole-shape matching demands the embedded figure be visible in the same orientation and size; the eye keeps re-scanning the same salient region because the target is treated as a gestalt rather than a set of line segments.
What to do instead
Before searching, reduce the target to one distinctive feature - the longest straight segment, an unusual acute angle, or a short parallel pair - and hunt only for that feature, then verify the remaining sides. Also consciously compare orientation: the embedded figure may be rotated, so scan for the segment in any direction.
On Army Aviation Information items, the candidate answers from general knowledge or gut feel - for example assuming a helicopter in autorotation still produces engine torque, or picking a control-response answer that sounds mechanically plausible but is not how the system actually works.
The domain is fact-dense and unfamiliar; partial exposure to aviation vocabulary creates false fluency, and the candidate substitutes plausibility for a memorized relationship.
What to do instead
Study in linked pairs and cause-effect chains rather than isolated definitions (collective = blade pitch and lift; cyclic = tilt of the rotor disc and direction; pedals = tail rotor thrust and yaw; autorotation = no engine drive, rotor driven by upward airflow). For each fact, be able to say what happens if the input increases and what happens if it fails.
On Spatial Apperception, the candidate reports the aircraft banking the wrong way - saying 'right bank' when the horizon tilts so that the left side of the panel is high - or mirrors left/right whenever the view is from the cockpit rather than from outside.
The candidate is judging the picture as an outside observer looking at a photo of an aircraft, not from the pilot's seat; in a banked attitude the horizon line rotates opposite to the wing that is down, which inverts the intuitive reading.
What to do instead
Fix the frame explicitly every time: 'I am the pilot.' Then use the rule that the wing on the side where the horizon appears higher is the wing that is down. For pitch, ignore the tilt and read only where the horizon sits vertically relative to the panel - above the panel means nose down, below means nose up.
The candidate treats pitch and bank as one judgment, so on a climbing-turn item they pick an answer choice that captures the turn but drops the climb, choosing the option that gets the bank right and the nose attitude wrong.
The tilted horizon is visually dominant and grabs attention first; the vertical position of the horizon is a smaller cue and gets processed only if deliberately checked.
What to do instead
Answer every item as two separate sub-questions in a fixed order - first: is the nose up, down, or level (vertical horizon position); second: which wing is down (horizon tilt). Only then read the answer choices, and eliminate any choice that gets either sub-answer wrong.
On Reading Comprehension, the candidate selects the choice that matches their own knowledge or opinion about the topic, or a statement that is true in the real world but never asserted in the passage.
Familiar subject matter triggers prior-belief activation, and the recognition of a familiar-sounding phrase feels like comprehension; the passage is read once for gist and never re-checked against the options.
What to do instead
For each option you are tempted by, return to the passage and put your finger on the exact words that support it; if you cannot locate supporting text, eliminate it regardless of whether it is true. Treat the passage as the only permitted source of facts.
The candidate picks answer choices containing absolutes or sweeping scope - 'always', 'never', 'all countries', 'the primary cause' - or a detail that is mentioned but is not what the question asked about.
These options are written fluently and echo the passage's vocabulary, so they feel authoritative; under time pressure the candidate matches words rather than testing the logical strength of the claim.
What to do instead
Pre-read the question stem and note exactly what is being asked (main idea, specific detail, inference, author's attitude). Then strike any option whose scope exceeds the passage, and prefer the choice whose claim is bounded to what the text actually states.
On Math Skills, the candidate sets up a rate, percent, or work problem with the quantities inverted - dividing where multiplication is needed, or computing 15% of the wrong base - and does not notice because the answer appears among the choices.
Word problems are translated directly into operations without writing down what each number means, and unit labels are dropped, so an inverted ratio still yields a plausible-looking number.
What to do instead
Write the units next to every number before computing (miles per hour, dollars per item), set the expression up so that unwanted units cancel, and sanity-check the magnitude with a rough estimate first - if the exact answer is wildly off the estimate, re-read the setup rather than the arithmetic.
The candidate burns four or five minutes on a single multi-step arithmetic item, then has to rush the last third of the section and leaves items unanswered.
Sunk-cost thinking plus the assumption that every question deserves equal effort; there is no pre-set abandonment rule, so the decision to move on is made emotionally and too late.
What to do instead
Adopt a hard cap: if no clean solution path appears within roughly 60-90 seconds, mark the item, pick a placeholder answer, and move on; return only after every other item has been answered. Never leave a blank, since there is no penalty for a wrong answer.
On Mechanical Comprehension, the candidate answers from everyday intuition - for example saying the larger gear turns faster, that a longer lever requires more force, or that a pulley system with more supporting ropes is harder to pull - instead of applying the stated principle to the diagram as drawn.
Casual physical experience is imprecise and often inverted (a big gear feels slow but is mentally linked with 'big = more'), and the candidate does not pause to identify which quantity is conserved before choosing.
What to do instead
Before looking at the choices, name the governing relationship out loud: gear teeth engaged means the same tangential speed so the smaller gear spins faster; torque equals force times distance from the pivot, so a longer arm needs less force; each additional supporting rope strand shares the load. Then compute or compare directionally, and use the specific values and geometry shown rather than what the setup 'usually' looks like.