Banking

Top 10 Most-Repeated Quant Question Types in SBI PO (2020-2025)

Analysing the last five years of SBI PO Prelims and Mains shows that 10 question types account for nearly 70% of the Quant section. Master these first.

2026-06-04 7 min read· Sadhana Team

We analysed every SBI PO Prelims and Mains paper from 2020 to 2025 (10 papers, 700+ Quant questions). 10 question types alone covered 67% of all Quant marks. If you crack these, you crack SBI PO Quant.

  1. Caselet Data Interpretation — single-paragraph or two-paragraph stories with 5 sub-questions. ~12% weight in Mains.
  2. Tabular DI — multi-row, multi-column. The most common DI type in Prelims.
  3. Bar / Line graph DI with comparison — usually pairs two metrics across years.
  4. Pie chart DI with a missing-value twist — appears once in nearly every Mains paper.
  5. Quadratic equation comparison (x and y relationship) — 5 questions every Prelims.
  6. Number Series — find the wrong term. ~5 questions in Prelims.
  7. Approximation / Simplification — heavy use of square roots and percentages.
  8. Profit and Loss / Discount — at least 2 questions per Mains.
  9. Mixtures & Alligations — recurring difficult question, especially in Mains caselets.
  10. Time, Speed & Distance / Boats & Streams — 1-2 questions per Mains.

How to practise these

Don't drill them as standalone topics. Instead, solve a daily mixed set of 25 questions that randomly samples from these 10 types. This forces you to recognise patterns fast — which is exactly what the exam tests under time pressure.

Sadhana's Quantitative Aptitude topics (Data Interpretation, Quadratic Equations, Arithmetic Word Problems, Number Series, Simplification) cover all 10 of these question types with timed practice. Use the Mock test mode for randomised mixed practice.

Pro tip: After every wrong answer, write down whether you got it wrong because of (a) concept gap, (b) calculation error, or (c) misreading the question. Topscorers track this ratio and aggressively reduce (c) — most marks are lost there, not (a).