Data Interpretation — IBPS PO Quantitative Aptitude

Data Interpretation is not a separate topic you study after arithmetic — it is arithmetic, wearing a chart. Every DI question is a percentage change, a ratio, an average or a total. IBPS knows this: in Prelims, 10–15 of the 35 quant questions are DI; in the Mains "Data Analysis & Interpretation" section it climbs to 18–25 of 40. Master three things — reading the chart before the question, picking the right base, and approximating instead of computing — and DI stops being the section that eats your clock and becomes the section that clears your cut-off.

1. What IBPS actually asks

DI arrives as a set of 5 questions on one chart or table. The chart types, in order of frequency:

1. Table DI — the purest form; rows × columns of numbers. Fastest to read, most common.
2. Bar graph — single, grouped or stacked bars; read heights.
3. Line graph — trends over time; watch the axis scale.
4. Pie chart — parts of a whole; everything is a share of a total (often given separately).
5. Caselet DI — a paragraph of data with no chart. You extract the numbers yourself — Mains' favourite, because the extraction is half the difficulty.
6. Missing DI / Mixed DI — some cells are blank and a clue lets you back-fill them; or two chart types combined.

Each set mixes four question flavours: totals/averages, percentage/ratio comparisons, percentage change (growth), and combination ("the value in 2022 is what % of the average of 2021 and 2023"). All four run on the same toolkit.

2. Read the chart before you read the question

The single biggest time-saver: spend 20 seconds reading the chart's structure first — what each row/column/axis means, the units (₹ crore? thousands? %?), and whether a total is given or must be computed. Candidates who dive straight into Q1 re-read the chart five times (once per question); candidates who map it once read it once.

Note the base. Pie charts and many tables give percentages of a total that is stated separately ("Total students = 4,800; the pie shows the % in each stream"). Half of all DI errors are base errors — taking a percentage of the wrong whole. Circle the total the moment you see it.

3. The three-tool toolkit (this is 90% of DI)

Percentage as a fraction. DI is engineered for the fraction table: 12.5% = ⅛, 16.67% = ⅙, 37.5% = ⅜. "Find 62.5% of 3,200" is a 3-second $\frac{5}{8}\times 3200=2000$, not a multiplication. This one habit halves your DI time.

Ratio comparison without dividing. To compare $\frac{a}{b}$ and $\frac{c}{d}$, cross-multiply: $\frac{a}{b}>\frac{c}{d}⟺ad>bc$. "Which department has the higher pass %?" never needs two divisions — one cross-multiplication answers it.

Percentage change. From old value $O$ to new $N$: $\%\Delta =\frac{N-O}{O}\times 100$. The base is always the earlier value. Growth from 250 to 300 is $\frac{50}{250}=20\%$ — never divide by 300.

$\text{Average}=\frac{\text{sum}}{\text{count}},\%\Delta =\frac{N-O}{O}\times 100,\frac{a}{b}\gtrless \frac{c}{d}⟺ad\gtrless bc$

4. Approximation — the exam-hall superpower

Mains DI (and the "Approximation" question type) does not want exact answers — the options are far apart on purpose. Round aggressively:

• 7,986 → 8,000, 31.2% → 31%, ₹4,930 → ₹5,000. Compute with round numbers, then sanity-check the options.
• First-digit / order-of-magnitude filtering: if the answer must be "roughly 2,400", eliminate the 240 and 24,000 options before any arithmetic.
• Ballpark averages: the average of five numbers between 400 and 600 is ~500 before you add anything — enough to kill two options.

The skill DI tests is decision speed under approximation, not precision. A candidate computing 46.8% × 7,983 exactly has already lost to the one who did 47% × 8,000 = 3,760 and picked the nearest option.

5. Caselet DI — extract, then it's a table

A caselet gives data as prose: "In a school of 1,200 students, 45% opted for Science. Of the Science students, the ratio of boys to girls is 7:5…" Your job is to convert the paragraph into a mini-table before answering anything:

1. Underline every number and its label as you read.
2. Build the table/tree on paper (Science = 540 → boys 315, girls 225…).
3. Only then read the questions.

The difficulty is entirely in the extraction; once the table exists, the questions are ordinary DI. Never try to answer a caselet question straight from the prose.

6. Missing / mixed DI

Some cells are blank, and a linking clue fills them: "The total for 2023 equals the total for 2022" or "B's value is 20% more than A's". Find the one cell you can compute from a clue, fill it, and the rest cascades. Treat it as a small equation, not a lookup.

Mixed DI stacks two representations (a table and a pie for the same data) — usually so one supplies the total the other needs. Match the shared quantity across the two before starting.

7. Order of attempt under the sectional clock

• Do Table DI first — least reading, fastest marks. Then bar/line, then pie, then caselet/missing.
• Skim the 5 questions of a set and do the "total/average" ones first (they need the whole chart anyway) before the fiddly "combination" one.
• One set, all five questions — the chart is already in your head; never solve a DI set for one question and leave.
• Budget 3.5–4.5 minutes per set. A set where you're re-deriving the base for the third time is a set to flag.

8. Solved set — a table DI

A shop sells four brands (P, Q, R, S). The table gives units sold and the selling price per unit:

Q1. Total revenue from R? $150\times 200=₹30,000$.

Q2. Q's revenue is what % of S's? Q $=80\times 400=32,000$; S $=90\times 500=45,000$. Ratio $=\frac{32000}{45000}=\frac{32}{45}\approx 71.1\%$. (Cross-check the options; ~71% wins — no long division needed.)

Q3. Average units sold across the four brands? $\frac{120+80+150+90}{4}=\frac{440}{4}=110$.

Q4. By what % do P's units exceed Q's? $\frac{120-80}{80}\times 100=\frac{40}{80}\times 100=50\%$. (Base is Q, the smaller/earlier value.)

Five questions, one table, under four minutes — the DI pattern in miniature.

9. The DI protocol

For every DI set, in order:

1. Read the chart structure and units (20 sec) — map rows, columns, axes, and circle any given total.
2. Skim all 5 questions; note which need the whole chart (totals/averages) vs one cell.
3. Convert every % to a fraction and approximate — round before you multiply.
4. Cross-multiply to compare ratios; never do two divisions.
5. Answer totals/averages first, then comparisons, then the combination question.
6. Sweep all five off the one chart; flag and leave any set past ~4.5 minutes.