Reasoning Ability — Previous Year Question Patterns

Representative question patterns and strategic notes reflecting recurring banking/SSC reasoning question types — illustrating the kind of applied technique previous-year-paper practice reinforces, not a substitute for solving the official previous year papers directly.

Q1 (Puzzle pattern — the "most definite clue" test). A circular seating puzzle includes both "P sits exactly opposite Q" and "R sits somewhere to the left of S, not necessarily adjacent." Which clue should be applied first, and why?

"P sits exactly opposite Q" — this is a definite, fixed positional relationship in a circular arrangement, immediately establishing two concrete positions. "R sits somewhere to the left of S, not necessarily adjacent" is a vaguer, less-constraining relative clue that narrows possibilities only once other positions are already established — applying it first, with no existing positions to narrow against, provides little immediate structural progress.

Q2 (Syllogism pattern — the "every diagram" test). Statements: "Some pens are books. All books are papers." Conclusion: "Some pens are papers." Is this conclusion valid, and what's the reasoning?

Valid. In every possible diagram consistent with both statements, the overlap between Pens and Books (from "some pens are books") must fall entirely within Papers (since "all books are papers" places the entire Books circle inside Papers) — meaning that overlapping region of Pens is necessarily also within Papers in every valid diagram, making "some pens are papers" true across all consistent diagrams, not just one specific drawing.

Q3 (Blood relations pattern — tree construction under ambiguous gender). "A is B's parent. B is C's sibling. D is C's child." Why is drawing the tree still necessary even though this chain seems short?

Even a seemingly short chain benefits from the tree specifically because "parent," "sibling," and "child" here are gender-ambiguous — the tree structure keeps the generational relationships (A's generation, B/C's generation, D's generation) visually clear without requiring you to resolve or assume genders that aren't stated, which is a common source of error when relations are tracked mentally and a candidate unconsciously assumes a specific gender not actually given in the statement.

Q4 (Direction sense pattern — a common trap). "A walks 4km east, then 3km north, then turns to face his starting point and walks toward it. In which direction is A now facing during this final walk?" What's the risk of solving this without a sketch?

Without a sketch, correctly determining "the direction facing the starting point" from a diagonal displacement (4km east + 3km north creates a diagonal offset, not a simple cardinal direction) requires accurately visualizing the resultant vector's direction, which is genuinely difficult to do reliably in one's head. A sketch on a simple grid, plotting the actual path and the resulting displacement, makes the final direction (southwest, in this case) visually obvious rather than requiring mental spatial reasoning under time pressure.

Q5 (Input-output pattern — full-step rule confirmation). Given three transformation steps in an input-output problem, a candidate identifies a rule that correctly explains step 1→2 and 2→3, and confidently applies it to answer a question about step 3→4 (not shown in the given steps). What's the risk in this approach?

The rule was only confirmed against 2 transitions (1→2 and 2→3) — if the actual underlying rule has a subtlety that doesn't manifest until a later step (for example, a rule that behaves differently once numbers exceed a certain value, or once all words have been moved), extrapolating to an unshown step based on a partially-confirmed rule risks applying an incorrect or incomplete rule. The safer approach is only extrapolating with high confidence when the given steps provide enough transitions to fully confirm the rule's consistent behavior, and treating an extrapolation based on minimal confirmation as lower-confidence.

Q6 (Inequality pattern — mixed chain with "either-or"). Statements: "P > Q, Q ≥ R." Conclusions: "I. P > R. II. P = R." Which conclusion(s) follow?

Only Conclusion I follows definitively — P > Q combined with Q ≥ R guarantees P > R (since P is strictly greater than Q, and Q is at least equal to R, P must be strictly greater than R in every case). Conclusion II (P = R) is never guaranteed by this chain, since P is always strictly greater than Q which is always at least R — there's no scenario consistent with the statements where P actually equals R. This tests whether a candidate correctly propagates a strict ">" through a chain that includes a "≥," which is a common site of mixed-chain errors.