BoM Generalist Officer (Scale II/III) common mistakes
BoM Generalist Officer (Scale II/III) · Understand the traps and how to avoid them.
In Reasoning puzzles and seating arrangements, the candidate fixes the first workable arrangement they find and stops, without testing whether the clues admit a second valid arrangement before answering the 'which is definitely true' question.
Under time pressure the relief of finding any consistent arrangement is mistaken for having solved the question; candidates also fail to notice 'definitely/possibly' wording, so a possible answer is accepted where a certain one is required.
What to do instead
Before answering, deliberately build the alternative: take the most loosely constrained clue and flip it, then check all clues again. If both arrangements survive, any option that differs between them is only 'possible', not 'definite'. Train by solving 20 puzzles where you must explicitly write 'Arrangement 1 / Arrangement 2' before choosing.
In syllogism and inequality (coded inequality) questions, the candidate judges validity by real-world meaning or by the direction of the symbol rather than by the formal rule for combining statements.
School logic taught 'all A are B' intuitively, and the exam's 'possibility' cases (some A being B is possible) conflict with that intuition; candidates also skip the step of chaining conclusions through a common middle term.
What to do instead
Adopt a mechanical method: draw the chain, apply only the standard rules (A>B, B>C gives A>C; A>B, B<C gives no relation), and treat 'possibility' as valid unless a definite opposite relation is forced. Practise 30 questions writing the chain in symbols before evaluating options, never reading the sentence as English.
In Data Interpretation and arithmetic, the candidate computes exact values for every sub-question instead of using approximation, ratio comparison, or percentage-change shortcuts, and runs out of time with 8-10 questions unattempted.
School practice rewarded exact calculation and full written working; the exam rewards selection. Candidates also re-read the chart for each sub-question instead of extracting all data once.
What to do instead
First pass: read the chart and note the base values, scale, and units. Second pass: for each sub-question decide whether approximation suffices (options far apart means estimate), and use fractions and percentage shortcuts. Do timed sets of 5 questions in 6 minutes with a calculator ban to build the estimation habit.
In arithmetic word problems (time-work, pipes, trains, mixtures, simple/compound interest), the candidate sets up the equation with the wrong base - e.g. uses total work as 1 but then adds rates of people who joined later, or applies simple interest for part of the period in a compound interest question.
The candidate memorises formulas as strings rather than as relationships, so the units and the 'per what' of each rate are lost; multi-stage problems (a pipe closed after some time) are read as single-stage.
What to do instead
Before calculating, write one line stating what the unit rate means ('A does 1/12 of the work per day') and list the timeline of who is active when. For interest, mark the compounding frequency and the exact period. Solve 15 mixed multi-stage problems where the timeline diagram is mandatory.
In English, the candidate answers Reading Comprehension and cloze from general knowledge or personal opinion instead of from the passage, and chooses an inference option that is true in the world but not supported by the text.
Familiar topics (banking, economy) trigger prior knowledge, and 'inference' is confused with 'plausible'. Candidates also read the questions after the passage and then hunt line by line, losing the argument's structure.
What to do instead
Read the passage once for structure (what is claimed, what is contrasted, what the author concludes), then answer with the rule: every chosen option must be traceable to a specific sentence or to a direct logical step from it. For inference questions, reject options that need outside facts. Practise 10 passages marking the supporting line for each answer.
In English error-spotting and sentence-improvement items, the candidate marks the part that 'sounds odd' rather than applying grammar rules, missing subject-verb agreement across an interrupting phrase, wrong prepositions, and faulty parallelism.
Spoken fluency masks written errors; the ear accepts 'one of the officers who was present' because it sounds smooth, and preposition collocations (comply with, consist of, prefer to) are learnt loosely.
What to do instead
Check in a fixed order: subject-verb agreement (strip the phrase between), pronoun reference, tense consistency, preposition collocation, parallelism in lists, and comparison completeness. Maintain a personal list of 50 collocations and revise it weekly; do 20 error-spotting items applying the checklist aloud.
In General/Banking Awareness, the candidate relies on stale static facts (older capital figures, former governors, superseded schemes) and cannot place current-events items because they never tracked the regulator's own releases.
Preparation is done from undated PDF compilations; the candidate never checks the as-of date, and exam questions are drawn from the last 6-12 months of RBI notifications, budgets, and policy rates.
What to do instead
Keep a one-page dated log updated monthly: policy repo rate and stance, CRR/SLR, key RBI circulars, new schemes and their launch amounts, heads of RBI/SEBI/NABARD/SIDBI, and budget allocations for banking. Before the exam, verify every figure against the primary source and discard anything undated.
In Banking and Financial Sector questions, the candidate treats similar terms as interchangeable - confuses CRR with SLR, NPA with restructured standard asset, repo with reverse repo, or Basel III capital buffers with the CRAR minimum - and loses marks on definitional items.
Terms are memorised as abbreviations without the mechanism: what is kept with RBI versus invested in approved securities, who borrows from whom in a repo, and which ratio includes which capital tier.
What to do instead
Build a comparison table with columns: definition, who holds the funds, instrument, effect on liquidity, and current value. For each pair, write one sentence of contrast in your own words. Test yourself by explaining CRR versus SLR and repo versus reverse repo to someone else without notes.
In Credit questions, the candidate applies thumb rules (e.g. a fixed debt-service coverage number or a standard margin) to every borrower without adjusting for sector, cash-flow seasonality, or facility type, and cannot compute or interpret DSCR, drawing power, or MPBF correctly.
Formulas are learnt in isolation from the lending decision; candidates do not practise the underlying cash-flow logic, so they mis-handle working capital assessment methods and confuse fund-based with non-fund-based exposure.
What to do instead
Work through 10 full credit cases: build the cash-flow statement, compute DSCR as (net profit + depreciation + interest on term loan) / (interest + term instalments), and derive drawing power from paid stock and creditors. State explicitly which assessment method applies and why, and note how a change in receivables days alters the requirement.
In Risk and Trade Finance, the candidate answers from generic definitions and misses the operational specifics - e.g. cannot distinguish a letter of credit from a bank guarantee, mis-states which party bears which risk, or fails to identify the risk type (credit, market, operational, liquidity, country) in a short scenario.
Trade products are studied as vocabulary rather than as flows of documents, goods, and money; risk categories are memorised as lists without the trigger conditions that make one dominant.
What to do instead
Draw the documentary flow for LC (sight and usance), bank guarantee, and bill discounting, marking at each step who is exposed and to what. For scenario questions, name the loss event first, then map it to the risk category and the mitigating control. Do 15 scenario items writing the risk type and one control before checking the answer.