HK Police RPC Written Exam common mistakes
HK Police RPC Written Exam · Understand the traps and how to avoid them.
In Chinese reading comprehension, answering from general knowledge or personal opinion about the topic instead of locating the statement in the passage that supports the chosen option, so answers reflect what the candidate believes rather than what the text says.
Candidates trained on current-affairs style questions assume the passage is background and the answer comes from outside knowledge; under time pressure they pick the option that 'sounds right' and never re-read the relevant lines, and Chinese passages on police-related topics invite this because candidates have prior views.
What to do instead
Adopt a locate-and-match routine: read the question stem first, identify the key term, scan the passage for the sentence containing that term, then choose only the option paraphraseable from that sentence. For inference items, mark the two options that go beyond the text and eliminate them explicitly. Practise with passages on unfamiliar topics so outside knowledge cannot rescue you.
Losing marks on Chinese grammar and written expression items involving measure words, aspect markers, and function-word pairs (e.g. 因為…所以, 雖然…但是, 不但…而且), and on register: writing colloquial Cantonese-style phrasing in a formal written Chinese answer.
Candidates who speak Cantonese daily rely on spoken intuition, and Hong Kong schooling often emphasises content over explicit grammar terminology, so the written-standard distinctions (書面語 vs 口語, 的地得 usage, 把/被 constructions) are never systematically drilled.
What to do instead
Keep a one-page error log of every grammar item missed, tagged by category (measure word, aspect, conjunction pair, 書面語 substitution). Memorise the fixed conjunction pairs and the common Cantonese-to-written-Chinese substitutions (e.g. 唔→不, 嘅→的, 佢→他). For writing tasks, draft then do a dedicated register pass replacing colloquial items before checking content.
In English, choosing an answer because it is grammatically plausible in isolation rather than because it fits the whole sentence or passage, and misjudging connector logic (however, therefore, despite, unless) in cloze and summary items.
Candidates read only the blank's immediate neighbourhood, and connectors are processed as vocabulary rather than as logical relations, so a sentence like 'Despite the delay, the unit proceeded' is misread as cause-and-effect. Time pressure reinforces single-sentence reading.
What to do instead
For every cloze blank, read the full sentence plus the preceding one and state the logical relation in your own words before looking at options (contrast, result, condition, addition). Eliminate options that are grammatically fine but logically wrong. Build a connector table mapping each connector to its relation and test yourself on mixed sets.
In English writing, producing an answer that is on topic but structurally loose: no clear topic sentence per paragraph, mixing formal report conventions with informal opinion, and inconsistent tense or pronoun reference across sentences.
Candidates prepare vocabulary and ideas but never rehearse a fixed structure, so under exam conditions they write in the order ideas arrive. Police-context writing (incident summaries, reports) rewards chronological and factual precision, which is different from the discursive style many candidates default to.
What to do instead
Memorise two templates: a factual report frame (time, place, persons, sequence of events, action taken) and an argument frame (position, two supports, concession, conclusion). Before writing, spend two minutes on a skeleton of topic sentences only. After writing, do one pass checking tense consistency and that every pronoun has an unambiguous antecedent.
In quantitative reasoning, setting up the correct calculation but losing the mark to unit and scale errors: mixing percentages and percentage points, treating a 20% increase as multiply-by-1.2 in one step but dividing by 0.8 when reversing it, or reading a chart axis that starts at a non-zero baseline.
Candidates focus on getting a formula and treat units as decoration; percentage reversal is a classic trap because 'increase by 20% then decrease by 20%' feels symmetric. Chart baselines are skimmed because candidates look at bar heights rather than axis labels.
What to do instead
Write the unit next to every number before calculating, and for any percentage change write the multiplier explicitly (increase 20% → ×1.2; reverse → ÷1.2, not ×0.8). For every chart, first state aloud the axis units, the baseline value, and what one gridline represents. Sanity-check the answer's order of magnitude against the raw figures.
In data interpretation and basic statistics, computing a mean, median, or rate from a table without checking whether the question asks about the whole population or a subgroup, and confusing 'average of the averages' with the overall average when group sizes differ.
Tables present rows and totals side by side, and candidates grab the most visually prominent row or the printed total without confirming it matches the question's population. Weighted-average problems look like simple-average problems until the group sizes are inspected.
What to do instead
Underline the population phrase in the question (all districts, only those aged 18–30, per 1,000 residents) and circle the matching column before touching numbers. When combining group averages, always multiply each average by its group size, sum, then divide by total size. Verify by checking whether the weighted result lies between the group averages.
In logical reasoning, accepting an argument's conclusion because it is plausible or matches the candidate's own view, rather than testing whether the stated premises actually support it, and treating a sufficient condition as necessary (affirming the consequent).
Candidates read arguments for content agreement rather than structure, and conditional reasoning is rarely formalised in everyday reading, so 'if P then Q; Q, therefore P' feels valid. Familiar police or law topics make the conclusion seem obviously true, masking the flaw.
What to do instead
For every argument, write the premises and conclusion as separate labelled lines before evaluating, and ask only 'do the premises guarantee the conclusion?'. For conditionals, diagram with arrows and test the contrapositive; memorise that P→Q licenses not-Q→not-P but never Q→P. On weaken/strengthen items, predict the answer before reading the options.
On Basic Law and National Security Law items, answering from general impressions of what the law 'should' say or from news headlines, and mixing up which instrument or article governs a point (e.g. attributing a provision to the wrong chapter, or confusing the Basic Law's amendment procedure with ordinary legislation).
These topics are encountered mainly through media coverage, which reports outcomes rather than provisions; candidates therefore hold confident but imprecise memories and do not distinguish between the Basic Law, the National Security Law, and ordinary ordinances.
What to do instead
Build a provision-level table with three columns: instrument, article number, and a one-line paraphrase in your own words. Drill only from the official texts and avoid secondary summaries for article numbers. For any question, first name the instrument and article before choosing, and mark items where you cannot do so for review rather than guessing.
On Hong Kong current affairs and government structure, relying on outdated knowledge of officeholders, bureaux, and statutory bodies, and confusing the roles of bodies with similar names (e.g. which body advises versus which decides, or which is a statutory body versus an advisory committee).
Candidates revise from notes made months earlier and from social media, and government reorganisation changes bureau names and responsibilities; similar-sounding councils and panels are memorised as a list of names without their functions.
What to do instead
Maintain a living one-page organogram of the key institutions with role, head, and whether statutory or advisory, and date-stamp every entry. Refresh from official government sources monthly in the run-up to the exam. When revising, learn each body by its function sentence rather than its name alone, and self-test by describing the function and naming the body.
In situational judgement, choosing the option that produces the most helpful or sympathetic outcome for the person in the scenario, or the most assertive action, instead of the option that upholds integrity, lawful procedure, and the chain of command; also failing to escalate when a matter exceeds the candidate's authority.
Candidates answer as a helpful member of the public rather than as a disciplined officer, and the 'nice' option is deliberately attractive. Integrity scenarios are missed because candidates assume misconduct must be blatant, overlooking procedural shortcuts, unauthorised disclosure, and acceptance of small favours.
What to do instead
Apply a fixed four-question filter in order: is it lawful, is it within my authority, does it protect integrity and confidentiality, and have I recorded and reported it? Eliminate any option involving unauthorised disclosure, personal benefit, or bypassing a supervisor, however helpful it looks. Where a scenario involves a colleague's misconduct, the correct answer almost always includes reporting through the proper channel, not informal handling.