Data Interpretation — DI in exam language — is exactly what it sounds like: you are given a chunk of data (a table, a bar graph, a line graph, a pie chart, or some combination), and then asked 4-5 questions about it. The data does not change between questions. You read once, answer five times.
Here is the analogy that makes this click: imagine you are a constable reviewing a register of duty hours at your police station. The register has columns — date, officer name, hours logged. Your senior asks you: "Who worked the most last week? What is the average for the month? What percentage more did Officer A work compared to Officer B?" You are not solving abstract maths — you are reading a structured record and computing relationships between numbers already in front of you.
That is all DI is.
The reason students lose marks in DI is not that the maths is hard. The underlying calculations are almost always percentage change, ratio, average, or simple arithmetic. The reason marks are lost is:
Your job in DI is to be a fast, accurate reader first, and a calculator second. The maths serves the reading — not the other way around.
SSC GD DI sets are typically based on one of four formats: tables, bar graphs, line graphs, and pie charts. Sometimes two formats are combined (a table with a graph alongside it). At the 10th-pass level, the data values are usually whole numbers or clean decimals, and the questions rarely require more than two or three arithmetic steps.
Tables: Rows and columns. Before you read any question, scan the headers — what is in each row, what is in each column, what is the unit (lakhs? thousands? percentage?). The unit row at the top is the single most-ignored line in a DI table. Ignoring it is how you pick "549000" when the answer is "990000".
Bar Graphs: Each bar represents one category or one time period. The y-axis is always the key — read its scale markings carefully. If the y-axis goes 0, 20, 40, 60 and a bar reaches halfway between 40 and 60, that bar = 50, not 45 or 55. Do not eyeball — find the nearest gridline and count.
Line Graphs: Used for data over time. Questions almost always ask for the year of maximum increase/decrease, or the percentage change between two specific years. The steepest slope = largest change. Use this visually before computing anything.
Pie Charts: Sectors represent parts of a whole, usually given as percentages or degrees. Key fact: percentage = (degrees / 360) × 100. If a sector is 72°, it represents 20% of the total. Questions often give you the total and ask for the value of one sector — multiply percentage by total. Or give you two sectors and ask for a ratio — you can often cancel the total out entirely.
1. Percentage Change
% change = [(New Value − Old Value) / Old Value] × 100
In DI, you are comparing two adjacent cells, bars, or points on a line. The question of "which period has the least percentage increase" is asking you to compute this ratio for every candidate period and pick the smallest. Do not compute all of them. Eliminate visually first — look for the period where the absolute jump is small relative to the base. Then verify with approximate arithmetic.
2. Ratio and Proportion
When a question asks "what is the proportion of A to B", you read both values and simplify. If Government owners = 1300 and Private owners = 300, the ratio is 13:3. No division needed — just find the GCD and divide. If the numbers are large, check the answer options first — they often tell you the simplified form, so you just need to verify which one matches.
3. Average
Average = Sum of all values / Number of values
In DI, you are almost always averaging over years (3 years, 5 years, 6 years). SSC GD will sometimes give you a set with 6 years of data and ask for the 3-year average of the first half vs. the last half. Add each group separately, then compare — you do not need to divide if you are only comparing which average is larger (just compare the sums).
Step 1 — 30 seconds: Read the title, all axis labels, all column/row headers, and the unit. Do not skip this.
Step 2 — 20 seconds: Skim the question stems (not the options) of all 4-5 questions. Identify which questions need exact calculation and which need approximation or visual reading.
Step 3 — Do the visual/approximation questions first. These give you 1 mark for essentially no arithmetic.
Step 4 — Do the calculation-heavy questions last. By now you have already read the data carefully, so you are not re-reading from scratch.
Step 5 — Always check units in your answer against the answer options before marking. If options are in thousands but you computed in raw numbers, you will pick the wrong bubble even with correct arithmetic.
SSC GD does not require exact computation in most DI questions. The options are usually spread far enough apart that rounding to the nearest 5% or 10% tells you the answer. If the question asks "approximately what percentage", that word "approximately" is a green signal — use estimation, do not long-divide.
Before computing any percentage change for a line graph, visually mark the two points with the steepest rise and steepest fall. Only calculate those two — ignore the rest. In a 5-period graph, this reduces computation from 4 comparisons to 2 at most. Standard method: compute all 4 changes (about 80 seconds). Slope scan + verify: 25 seconds. Works because "highest percentage change" almost always corresponds to the steepest visible slope when the x-axis intervals are equal.
When a question asks for a ratio (Government : Private, Company A : Company B), look at the answer options before you compute. If options are 13:4, 13:3, 3:13, 4:13 — you only need to determine two things: (a) is the first number larger or smaller? (this eliminates 2 options immediately) and (b) is the second number 3 or 4? (one subtraction or approximation confirms this). Standard method: compute both values and simplify (45 seconds). Options-first elimination: 15 seconds. Works for ratio questions where options give you the simplified form.
When the question asks which of two groups has the higher average, and both groups have the same count of items, just compare their sums — do not divide. If Company CTU's first-3-year total is 180 and last-3-year total is 210, the last three years have a higher average. No division needed. Saves 2-3 steps. Standard method: compute both averages (30 seconds). Sum comparison: 10 seconds. Valid only when the number of items in each group is identical.
If a pie chart gives sector percentages and you need the actual value of one sector: Sector value = (% / 100) × Total. Do the multiplication, not a proportion cross-multiply. If private owners are 5% of 1,10,000 total: 5/100 × 1,10,000 = 5,500. Direct multiplication: 2 steps. Cross-multiplication: 4 steps. This also works in reverse — if you know sector value and total, % = (value/total) × 100. Keep this direction clear or you will invert the fraction.
For "which period has the least/most percentage increase" questions, do a quick visual scan. The least percentage increase occurs when either (a) the absolute increase is small, or (b) the base (previous year value) is large. Scan both conditions visually — periods with a tiny bar/line jump over a large base are your candidates. Compute only those 1-2 candidates precisely. Standard method: compute % change for all periods (5 periods × 15 seconds = 75 seconds). Visual elimination + compute 1-2: 20-25 seconds.
When you open a DI set in the exam hall, do this in order:
Read first (50 seconds total): Title → Units → All headers/labels → Skim all question stems.
Classify each question:
Attack order: Ratio and visual questions first. Percentage-of-total questions second. "Which period has highest/lowest change" questions third (these need comparison across multiple periods). Exact large-number computation questions last.
Unit check: Before marking any answer, confirm your answer's unit matches the answer options. If options are in "thousands" and you worked in raw numbers, shift your decimal.
Do not recalculate a value you already found. Write the two or three key values you read from the data at the top of your rough work. You will need them across multiple questions in the same set.
Why this question: This tests whether you can compare percentage changes across multiple periods without computing all of them — a core DI skill.
Solving path: For each period, compute % increase = (increase in that period / total at start of period) × 100. The period where this ratio is smallest is your answer. Use the visual elimination trick — look for the period where the jump in internet owners is small relative to the existing base. For 2001-02, the absolute increase is relatively small compared to the large base built up by then, making the percentage increase the least. Verify with approximate arithmetic to confirm 2001-02 before marking.
Why this question: Tests exact reading of "fresh owners" — a derived value (current total minus previous total). Common trap: students read the total for 2001-02 instead of computing the increment.
Solving path: Fresh internet owners in 2001-02 = (Total in 2001-02) − (Total in 2000-01). Read both values from the data carefully, paying attention to units (the answer 990000 vs 99000 is entirely a units question). Subtract: the difference is 990000. If you got 99000, you read one of the values in the wrong unit scale.
Why this question: Tests ratio reading with a direction trap — the options include both 13:3 and 3:13, so you must get the order right.
Solving path: Question asks Government : Private. Read Government owners and Private owners for 1999-2000 from the data. The question asks "proportion of Government to Private" — Government comes first. If Government = 1300 (units) and Private = 300 (units), ratio = 1300:300 = 13:3. The trap option 3:13 would be the answer if you reversed the order. Always re-read "A to B" — A is the numerator.
Why this question: Combines two operations — find Private owners in two different periods, then compute percentage increase between them. Tests whether you can chain operations without losing track of which value is the base.
Solving path: Step 1 — read Private owners in 2001-02 (call it P2). Step 2 — read Private owners in 1998-99 (call it P1). Step 3 — % increase = [(P2 − P1) / P1] × 100. The answer is approximately 8000%. Note: an 8000% increase sounds enormous, but if P1 was very small (say, 10,000) and P2 is large (say, 8,10,000), the arithmetic works out. Do not second-guess a large percentage answer in DI — verify the arithmetic, then trust it.
Why this question: Tests percentage-of-total reading. One of the most frequently repeated DI question types in SSC GD.
Solving path: % of Private = (Private owners in 1998-99 / Total owners in 1998-99) × 100. Read both values. If Private = X and Total = Y, compute X/Y × 100. The answer is approximately 5. Use the options to guide your estimation — if you are getting something between 4 and 6, pick 5. If your rough calculation gives 18-22, you may have swapped numerator and denominator (computed Total/Private instead of Private/Total).
Ignoring the unit row. The difference between "990000" and "99000" in a question is entirely about whether you registered that the data is in thousands or raw numbers. Read the unit before you read a single data value.
Reversing ratio order. "Proportion of A to B" means A:B, so A is the numerator. If the question says "Government to Private", Government goes first. The trap options always include the reversed ratio — eliminate the reversed one immediately by checking which number should be larger.
Using total instead of increment for "fresh" or "new" questions. "Fresh owners" in a period = current cumulative total minus previous cumulative total. It is not the total for that period. This trap appeared directly in the 2023 PYQ set.
Computing when comparison was enough. For "which year has the highest average", compare sums when group sizes are equal. For "which period has the steepest rise", scan visually before computing. Students who compute everything run out of time.
Misidentifying the base year in percentage change. % change from A to B = [(B − A) / A] × 100. The base is always the earlier/starting value. If you divide by B instead of A, you get a different (wrong) answer — and one of the wrong options is often designed to match that mistake.
Not reading all question stems before starting. Two questions in the same set often use the same intermediate value (say, total owners in 1999-2000). If you compute it once for Q1 and write it down, you save recalculation time for Q3. Students who read one question at a time repeat work needlessly.