UM UGM CBT common mistakes
UM UGM CBT · Understand the traps and how to avoid them.
Spending near-equal time on every Tes Potensi Skolastik subtest item, so a single long verbal-analogy or number-sequence item eats 3–4 minutes and forces rushed guessing on the last 10 questions of the section.
Candidates treat the TPS as a knowledge test where finishing every item correctly matters, when it is a speed-and-prioritisation test with a fixed per-section clock; the habit comes from school exams where unfinished items are rare.
What to do instead
Before the exam, set a personal cap per item (roughly 60–75 seconds on TPS verbal/numerical items). In practice sets, use a visible timer and physically skip any item past the cap, marking it for a second pass. Train the skip reflex until it is automatic, and rehearse a 2-minute end-of-section sweep for marked items only.
Reading the whole passage first and then answering Literasi questions from memory, without returning to the text, so answers reflect general knowledge or personal opinion rather than what the passage states.
School reading habits reward comprehension and recall; candidates assume understanding the gist is enough, and under time pressure re-reading feels wasteful.
What to do instead
Adopt a question-first or paragraph-mapping routine: skim the questions, then read paragraph by paragraph while tagging each paragraph's function (claim, evidence, counterargument, conclusion). For every answer, point to the exact sentence that supports it before selecting. Practise with paired passages where the correct answer hinges on one qualifying word.
Choosing a Literasi answer that is true in the real world but not supported by the passage, or that overstates the author's claim (e.g., picking 'all' when the text says 'many').
Candidates import outside knowledge and are attracted to strong, confident-sounding options; they also conflate 'the author would agree' with 'this is a reasonable general statement'.
What to do instead
Apply an explicit two-filter check on every option: (1) Is this stated or directly implied in the text? (2) Does the scope match exactly - same quantifier, same strength of verb? Eliminate options that add causes, comparisons, or absolutes absent from the passage, even if they sound sensible.
Setting up Penalaran Matematika word problems by grabbing numbers and computing immediately, without defining the variable or writing the relationship, then discovering the computed quantity is not what was asked.
Time pressure and drilled arithmetic fluency push candidates toward calculation; Indonesian word problems often bury the actual question in the final clause and include distractor data.
What to do instead
Force a 10-second written plan: underline what is asked, list given quantities with units, and write one equation or ratio before touching the calculator. After computing, restate the answer in the units and form the question demands. Drill 20 mixed word problems where at least three data points are irrelevant, to build the habit of filtering.
Losing points on Penalaran Matematika because of unit and scale slips - mixing percent with percentage points, treating a 20% discount followed by 10% as 30%, or ignoring that a chart axis starts at a non-zero value.
These items deliberately test proportional reasoning, and candidates apply additive intuition to multiplicative situations; chart axes and unit labels are skimmed rather than read.
What to do instead
Build a one-page error log of these traps and rehearse the correct model: successive percentage changes multiply (0.8 × 0.9), percentage-point differences are not relative changes, and each graph must be checked for axis origin and units. Re-solve every missed item the next day and classify the error type.
Treating Mata Pelajaran Pendukung Program Studi as a repeat of the school subject syllabus and revising broad content instead of the specific higher-order reasoning the exam uses (e.g., analysing an experimental design in Biology for a Medicine-track candidate).
Candidates rely on school notes and rote definitions, but the subject-support section tests application, data interpretation, and causal reasoning within the subject, often through stimulus-based items.
What to do instead
Revise by topic using application-level question stems: for each core concept, practise explaining a graph, identifying a flawed conclusion, and predicting an outcome. Prioritise the topics most heavily weighted in the target programme's subject and do timed sets of stimulus-based items rather than definition flashcards.
Misallocating the whole test: burning too long on the first section and arriving at later sections with reduced focus, or leaving no buffer for transferring answers and checking the answer sheet.
Section timers are independent, so candidates assume leftover time carries over, and adrenaline in the first section encourages over-investment; fatigue later compounds it.
What to do instead
Rehearse full-length simulations with the real section order and breaks, logging time per section. Set a target finish time per section that leaves 2–3 minutes for answer-sheet verification, and practise a 30-second physical reset (shoulders, breathing, eyes off screen) between sections.
On TPS numerical and figural reasoning items, guessing from the first pattern noticed without testing it against all given terms, so a sequence rule that fits the first three numbers fails on the fourth.
Pattern-recognition is fast and satisfying, and candidates rarely verify because verification feels like wasted time; figural items also invite superficial rotation/reflection guesses.
What to do instead
Use a fixed verification step: derive the rule, then apply it forward to reproduce every given term before predicting the missing one. For figural series, track two attributes separately (element count and position/rotation) and check both. If two rules fit, prefer the simpler and confirm it predicts all terms.
Answering every item with no deliberate guessing policy, leaving blanks or spending time on items that cannot be solved, instead of eliminating options and committing to the best remaining choice.
Candidates carry over penalisation fears from other tests or hesitate because they want certainty; on a CBT with no negative marking for wrong answers, blanks are pure lost expected score.
What to do instead
Confirm the scoring rule for the exam and adopt a standing policy: eliminate at least two options, choose among the rest, and never leave an item blank. In practice, track how many items you left blank and treat any blank as an error to eliminate in the next simulation.
Practising only on paper or untimed PDFs, then struggling with the CBT interface - scrolling long stimulus text, using the on-screen question navigator, and managing the countdown clock visible on screen.
Content knowledge is over-practised relative to format familiarity, and the anxiety of a visible timer plus navigation controls consumes working memory that should go to the items.
What to do instead
Do at least three full simulations in a CBT-like environment: on-screen text, a visible countdown, and a question-jump panel. Learn the navigator so you can flag and return to items quickly, and practise reading long passages on screen without printing.