KRA CBT Aptitude Test common mistakes
KRA CBT Aptitude Test · Understand the traps and how to avoid them.
Spending equal time on every numerical item, so a long multi-step percentage/ratio problem consumes four minutes and forces rushed guesses on the last five questions.
Candidates treat the section as a test of correctness rather than of scoring rate; they have no pre-set time cap per item and no triage rule, so sunk-cost thinking keeps them on the first hard question.
What to do instead
Before practising, set a hard per-item budget (roughly 60-75 seconds) and rehearse a two-pass routine: answer all items solvable in under a minute, flag the rest, return only if time remains. In practice, log the time each item took alongside whether it was right, and drill the habit of abandoning an item at the budget mark even when the answer feels close.
Estimating where exact arithmetic is required, or computing exactly where approximation would do, on percentage-change, ratio and average questions.
Candidates lack a decision rule for when answer choices are far enough apart to permit rounding; they also mis-handle percentage change by applying successive changes additively or by computing change relative to the wrong base.
What to do instead
Build a habit of scanning the answer options before computing: if options are widely spaced, use rounded figures (e.g. treat 19.8% as 20%) and only refine if two options collide. Separately drill the base rule: percentage increase = change divided by the original value, and successive changes multiply (1.2 x 0.9), never add. Write the base value explicitly before each calculation.
Answering verbal questions from general knowledge or personal opinion rather than from statements in the passage, and choosing an option that is true in the world but not supported by the text.
Passages are often on familiar workplace or social topics, so the brain auto-fills conclusions; candidates also confuse 'inference' with 'plausible guess' and never return to locate the supporting line.
What to do instead
For every answer, point to the specific sentence or phrase in the passage that supports it, or reject it for having no anchor. Practise classifying options into supported, contradicted, and not-mentioned, and drill 'cannot be inferred' items separately, since these punish imported knowledge most heavily.
Missing negation, exception and qualifier words ('except', 'not', 'least', 'only if', 'unless') and selecting the true statement when the question asks for the false one.
Under time pressure readers skim for topic keywords and skip function words; the question stem is read once, quickly, before the options, so the polarity of the task is never encoded.
What to do instead
Adopt a fixed routine: read the stem first, underline or mentally capitalise the operative word (EXCEPT, NOT, FALSE), then read the options. At the end, re-read the stem and confirm your chosen option answers the question actually asked. Drill batches of mixed-polarity items until the check becomes automatic.
Treating each abstract figure as a new puzzle and searching for any pattern, rather than systematically checking element properties in a fixed order.
Candidates have no checklist, so they notice the most visually salient feature (often colour or size) and build a rule on one example, then fail when the rule does not hold across the series.
What to do instead
Use a fixed scan order for every item: number of elements, type of element, position/rotation, shading, size, then line count or intersections. Verify a hypothesised rule against at least two consecutive frames before applying it, and reject any rule that fits only one transition.
Assuming a sequence's rule continues in the same direction or arithmetic operation, when the series alternates, uses two interleaved sequences, or changes operation (add 3, add 6, add 12).
The first difference is computed and, once it looks regular, the candidate stops testing; alternating or multiplicative patterns are invisible if only consecutive differences are examined.
What to do instead
For number series, always compute first differences, then second differences, then check ratios; if neither is clean, split into odd and even positions and analyse each subsequence. Keep a written list of candidate operations rather than committing to the first one that produces a plausible next term.
Comparing strings, codes, dates or figures by overall impression instead of character by character, so a transposed pair or a single differing digit is missed.
Skilled readers process words and numbers as whole units, which is efficient for comprehension but actively suppresses detection of small internal differences; the eye also jumps rather than tracking every character.
What to do instead
For comparison items, physically point or use the cursor and compare left to right in fixed chunks (e.g. three characters at a time), and for numbers align digits by place value before comparing. Practise short daily accuracy sprints where you mark every difference in paired strings, scoring yourself on misses rather than speed.
Rushing the perceptual speed section and trading accuracy for volume, so a fast finish produces more wrong answers than a measured pass.
The section is explicitly timed and visually repetitive, which creates urgency and fatigue; candidates assume raw speed is scored without penalty for errors, and accuracy degrades sharply after the first minute of unbroken scanning.
What to do instead
Check the scoring rule first: where accuracy is penalised, target a sustainable rhythm rather than maximum speed, and deliberately slow on the first ten items to establish a clean method. Use a brief reset (look away, one deep breath) every 20-30 items to counter fatigue, and track accuracy percentage alongside items attempted in practice.
Reading a table or chart value off the wrong row, column, unit or axis, or ignoring the legend, footnotes and stated base year before calculating.
Layout-driven errors happen because candidates locate the number that looks right rather than tracing the row and column label; units (thousands vs millions) and percentage-vs-percentage-point distinctions are skimmed.
What to do instead
Before any calculation, state aloud or in writing what the axes, units and legend mean and which row/column you need. Trace with a finger or cursor from the label to the cell, and convert all figures to a single unit before computing. For change questions, note explicitly whether the answer is in percentage points or percent.
In situational judgement items, choosing the option that is most decisive, most compliant-sounding or most personally sympathetic, rather than the one matching the stated workplace context and escalation norms.
Candidates answer as themselves or as an idealised employee, ignoring the scenario's constraints (deadline, authority level, safety, confidentiality); they also pick extremes because strong-sounding answers feel more competent.
What to do instead
Read the scenario for explicit constraints and your role's authority, then rank options by professional priority order: safety and legality, then confidentiality and fairness, then task delivery, then relationship comfort. Prefer proportionate action (raise with the right person, document, follow procedure) over both silence and unilateral escalation, and practise explaining why each rejected option fails.