Psychometric Test (PET) common mistakes
Psychometric Test (PET) · Understand the traps and how to avoid them.
Treating a passage question as a memory test: reading the text once, then answering from general knowledge or from the first sentence that echoes a keyword in the question.
The verbal section is timed and passages are dense, so candidates rush the read and rely on gist. Distractor options are deliberately written to match surface wording, so a keyword match feels like confirmation.
What to do instead
Read the question stem before the passage so you know what to hunt for, then locate and underline the one or two lines that actually decide the answer. Reject any option you cannot tie to a specific line; if you cannot point to the line, mark the question and move on.
Confusing 'must be true' with 'could be true' in inference and assumption items, picking the strongest-sounding option rather than the one the text forces.
Everyday reading rewards plausible elaboration, and the strongest option often reads as the most confident and therefore correct. Candidates also import outside facts that are true in the world but absent from the passage.
What to do instead
Before answering, say the task word out loud in your head: 'must', 'could', 'weakens', 'assumption'. For must-be-true items, apply the denial test: if the option were false, would the passage contradict itself? If not, it is wrong. For assumption items, check that the option is necessary, not merely helpful.
Solving quantitative items by setting up full algebra when the answer choices allow back-solving, or conversely estimating when the question demands an exact value.
School training conditions candidates to 'show the work', and under time pressure the familiar algebraic route feels safer. Estimation errors then go unnoticed because no exact check is performed.
What to do instead
Scan the answer choices first. If they are numbers, plug them back in or work backwards from the middle option; if they are far apart, estimate to one significant figure and pick the nearest. Reserve full algebra for questions where the choices are close together or symbolic.
Losing track of units, percentages, and base values: adding a 20% increase to a 20% decrease and expecting the original number, or mixing rates per hour with rates per minute.
Percentage change and rate problems are written to exploit the intuition that changes are symmetric and that units can be carried loosely. Candidates also forget whether a percentage is of the original or the new value.
What to do instead
Write the unit next to every number as you copy it, and convert to a single unit before calculating. For successive percentage changes, use multipliers (1.2 × 0.8 = 0.96, a 4% net decrease) rather than adding percentages. Sanity-check direction: a rise then equal fall must end below the start.
Spending disproportionate time on one hard quantitative item while easier items later in the section go unanswered.
Sunk-cost pressure: having invested two minutes, candidates feel the item is nearly solved and refuse to abandon it. Sections are usually scored by number correct, so an unfinished easy item costs the same as a missed hard one.
What to do instead
Adopt a per-question time cap (roughly one minute for a 20-question, 20-minute section) and a hard stop rule: if no clean path appears within the cap, mark, guess, and move. Return only if time remains. Practise with a visible timer so the cap becomes automatic.
Describing a figural sequence item by what the shapes 'look like' rather than isolating one transformation at a time, so the rule is guessed from the whole panel at once.
The eye processes the panel holistically, and salient features (colour, size) dominate attention while the actual rule may be rotation of an inner element or a count of intersections. Candidates then invent a rule that fits two frames but not the third.
What to do instead
Track attributes separately in a small table: position, orientation, number of sides, shading, and count of elements. Verify a proposed rule against every frame, including the last before the blank. If it fails on any frame, discard it and test the next attribute.
Assuming a rotation is clockwise when the figure is symmetric, or missing that a shape has been reflected rather than rotated.
Symmetric figures look identical under several transformations, so a rotation and a reflection can produce the same appearance. Candidates commit to the first transformation that seems to work and stop checking.
What to do instead
Use an asymmetric marker inside the figure (a dot, a notch, a shaded corner) to track orientation. Test reflection explicitly by checking whether the marker's handedness has flipped. When two rules both fit, choose the one that also explains the distractor options.
Ignoring matrix and series direction conventions: reading rows left to right when the rule runs down columns, or treating a series as arithmetic when it alternates between two interleaved sequences.
Left-to-right reading is automatic, and single-rule expectations lead candidates to force one operation onto a sequence that actually combines two. Alternating series are a standard construction and look irregular until split.
What to do instead
For matrices, test row rules and column rules separately before combining them. For number or figure series, check alternate terms (1st, 3rd, 5th versus 2nd, 4th, 6th) and differences of differences before concluding the pattern is broken.
Answering English usage items by ear, choosing the option that 'sounds right' in casual speech, especially on subject-verb agreement with intervening phrases and on pronoun case after prepositions.
Spoken English tolerates constructions that written standard English rejects, and the ear is trained on speech. Long noun phrases between subject and verb also dilute the agreement cue.
What to do instead
Strip the sentence to its skeleton: delete prepositional phrases and relative clauses, then check agreement between the bare subject and verb. For pronoun case, remove the other party ('between you and me' becomes 'between me'). Verify each option against a stated rule, not a feeling.
Misreading the task in English proficiency items: writing a summary when asked to argue, exceeding or falling short of the required word count, or omitting the register the prompt specifies.
Candidates prepare generic essay templates and apply them regardless of the prompt. Task achievement and length limits are separately scored, so a well-written response on the wrong task loses heavily.
What to do instead
Underline the instruction verbs and the word limit before writing, and allocate two minutes to a bare plan that maps each required element to a paragraph. Count words at the end and adjust. Match register to the stated audience: no contractions or slang in formal tasks.