The Mains stage rewards integration: recognising a named distinction — CRR versus SLR, weighted versus simple average, inference versus restated detail, one payment rail versus another — inside differently worded questions. This guide builds those distinctions, works two scenarios where the obvious first step gives a wrong answer, and finishes with a five-week sequence scored against explicit readiness checks. Administrative details such as dates and structure should come from the recruitment notification on SBI's careers page.
Policy Tools You Must Keep Apart: CRR, SLR, Repo and Reverse Repo
Financial awareness questions separate CRR, SLR, repo and reverse repo by three questions: who acts, what moves, and where the money sits. Build a comparison table and recite it in both directions.
The repo rate is the price the Reserve Bank of India charges when it lends to banks against government securities, and the reverse repo is the mirror operation in which the RBI absorbs surplus liquidity by paying banks to park funds with it. The cash reserve ratio is a share of a bank's net demand and time liabilities held as cash balances with the RBI itself, while the statutory liquidity ratio is a share the bank holds in its own portfolio as liquid assets such as government securities, cash or gold.
Revise the table in both directions. Given a news line such as a rate decision, ask which tool was in play and what it controls; given a tool name, state its operation from memory. This both-directions drill turns four similar-sounding terms into four separate retrieval cues, and a single session takes only a few minutes, which makes it easy to repeat weekly alongside your awareness log.
| Tool | Type | What moves | Where the money sits |
|---|---|---|---|
| Cash reserve ratio (CRR) | Quantity tool | Liquidity available for lending | Cash balances with the RBI |
| Statutory liquidity ratio (SLR) | Quantity tool | Required liquid-asset holdings | The bank's own portfolio (government securities, cash, gold) |
| Repo rate | Price tool | Cost of RBI lending to banks | Collateral: government securities |
| Reverse repo | Price tool | Return on parking surplus funds | Bank deposits with the RBI |
Anchoring Current Affairs to Institutions, Schemes and Regulators
Treat every awareness news item as a bundle to unpack: the institution behind it, the scheme or product named, and the regulator or ministry responsible. Three tags per item beat one memorised headline.
Maintain a dated log and, for each item, record the actor (the RBI, a bank, a market regulator, a ministry), the object (a scheme, a rate decision, an appointment, a digital product), and one sentence on why that actor has authority over that object. An isolated fact such as a scheme launch fades quickly; a tagged item connects to everything else you know about the same institution, so revision compresses instead of growing.
Once a week, cover the log and reconstruct each item from its tags alone; anything you cannot reconstruct goes back on the revision list. Connect this to the policy-tool work in the first section: every rate-decision entry in the log should trigger you to state which tool was involved and what it controls, so awareness and banking concepts reinforce one another rather than competing for study time.
Caselet DI: The Weighted-Average Trap and How to Sidestep It
Caselet questions hide the base figures in sentences rather than charts. When a question asks for an average across groups, decide between a simple and a weighted average before touching the calculator.
Worked scenario: a caselet reports Branch A holding deposits of 40 crore with 25% in current accounts, and Branch B holding 10 crore with 60% in current accounts. Asked what share of the combined deposits is current-account, the tempting step is to average 25% and 60% and answer 42.5%. The better decision is to weight each percentage by its base: 10 crore plus 6 crore against 50 crore gives 32%. The mistake matters because a follow-up ranking question in the same caselet would flip with it, so one wrong decision compounds across the set.
Practical exercise: take any two-group caselet from a practice set, compute the simple average and the weighted average of the same percentage, and record how far apart they are. Repeat with three groups. Expected observation: the gap widens whenever group sizes are unequal, so from now on an unequal caselet should trigger a deliberate check of denominators — which percentage belongs to which base — before you commit to an answer.
Seating Puzzles: Test the Facing Assumption Before You Fix Positions
Conditional reasoning puzzles fail at the first clue when a facing direction is assumed instead of derived. Enumerate the facing cases explicitly, then let the remaining clues eliminate one.
Worked scenario: a puzzle has five people facing an unspecified direction around a table, and clue one reads 'P is to the immediate right of Q'. If everyone faces the centre, P sits on one side of Q; if everyone faces outward, the same words place P on the opposite side. Committing to one facing locks every later clue to that choice; one contradiction and the arrangement must be rebuilt. The better decision: draw both facing versions side by side and continue until a clue eliminates one. The discarded tree is cheap; the wrong commitment costs the puzzle.
Carry the same discipline to input-output questions, where the machine's rule is itself a conditional. Identify whether elements sort by value, alternate between two operations, or shift position one step per pass, and verify the candidate rule against the last given step before predicting anything. A rule that explains the second step but not the fourth was never the rule, and catching that in the verification step is what keeps a five-part question set intact instead of losing every part to one bad pattern.
English: Answering Inference Questions From Evidence, Not Association
Detail questions can be answered by matching words; inference questions require a step of reasoning the passage supports without stating. Classify each question before reading options, because the two need different verification.
Classify first. A detail question is settled by locating the sentence that restates the answer, so scan for the paraphrase of the stem. An inference question is settled by combining two statements, or a statement with the passage's tone; the correct option will be weaker and duller than the most vivid sentence, because a supported inference stays close to the text. An option that merely echoes the most vivid sentence is a restatement of detail, which an inference stem did not ask for — recognising that gap between echo and support is the core skill.
Verification drill: after answering a five-question passage set, mark each option as stated (S), implied (I), or neither (N) by the passage. If you marked an inference-question answer as S, you probably selected a restatement — reread the stem before moving on. Expected observation: two or three options per set fall into N, and they are usually the extreme or emotionally loaded ones, so learning to spot N-options at a glance is what makes the classification fast enough to use under time pressure.
Computer Aptitude: One Payment Rail at a Time
Digital literacy questions reward separating payment systems by settlement model and typical use rather than by name familiarity: NEFT batches, RTGS real-time, IMPS instant mobile, UPI virtual addresses.
Anchor each rail to its distinguishing mechanism. NEFT settles in periodic batches, so it suits transfers where a short delay is acceptable. RTGS settles each transaction individually and continuously, which is why it is positioned for high-value transfers. IMPS moves money instantly through mobile and branch channels around the clock, and UPI sits on top of the instant-payment infrastructure to let users pay with a virtual payment address instead of sharing account details. The settlement model — batch, continuous, or instant — is the axis on which one-rail-versus-another questions turn.
Then widen to the rest of computer aptitude with the same actor-object habit you use for current affairs: a question about a file format, a keyboard route, or a networking term (LAN versus WAN, a browser versus a search engine) is asking who or what acts on what. Build one-page distinction sheets for pairs that sound alike, and test yourself by writing the distinction from memory before rereading the sheet. Recognition feels like knowledge; writing it out is what proves the retrieval works.
A Five-Week Sequence with Readiness Checks You Can Score
Sequence the weeks by concept family rather than by subject label: tools and awareness first, quant and reasoning next, integration last. Score yourself against a written rubric, not a feeling.
Weeks one and two: build the distinction sheets — policy tools, payment rails, scheme tags — and drill one quant or reasoning concept family per day, ending each day with retrieval in both directions. Week three: mixed timed sets per subject, logging every mis-solved caselet or puzzle with the decision that went wrong, not just the answer. Week four: full mixed sittings, reviewing the decision log between them. Week five: light revision of your own sheets and the awareness log. Adapt the proportions to your available time; the order — distinctions, single-concept drilling, mixed sets, integration — is the transferable part.
Readiness checks, as learning milestones rather than predicted scores: reproduce the four-tool table unaided in both directions; compute a weighted average before a simple average comes to mind on unequal groups; draw both facing trees on an unspecified-direction puzzle as a reflex; classify an English set into detail versus inference with a written rationale per question; state each payment rail's settlement model from memory. An unmet check tells you which week of the sequence to revisit — that is what the rubric is for. Dates, structure and eligibility belong to the recruitment notification on SBI's careers page.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.