Data interpretation questions appear in virtually every numerical aptitude test. Whether you're taking SHL, Kenexa, or AMCAT, you'll face questions that require you to extract information from charts, graphs, and tables — then calculate an answer under time pressure.
The most common chart type in aptitude tests. Key things to check immediately: Does the y-axis start at zero? (If not, differences between bars are visually exaggerated.) What are the units? (Thousands? Millions? Percentages?) Are bars grouped or stacked? Grouped bars sit side by side; stacked bars require you to read segments within each bar.
Used to show trends over time. Key checks: What's the time interval on the x-axis? (Monthly? Quarterly? Annual?) Are there multiple lines? (If so, read the legend carefully.) What's the y-axis scale? Questions often ask about rates of change — the steepness of the line matters more than the absolute value.
Show proportions of a whole. Key checks: Is the total given? (You'll need it to calculate absolute values from percentages.) Are percentages shown or do you need to estimate from the visual? Pie chart questions often combine with a total figure: "If total revenue was £2.4M and Region A represents 35%, what was Region A's revenue?"
Raw data in rows and columns. Tables are the most information-dense format. Key checks: Read ALL column headers before looking at the data. Check for footnotes — they often contain critical information like "figures in thousands" or "excludes VAT." Identify which row and column the question refers to before calculating.
Step 1: Read the question first. Before looking at the chart, read the question. "What was the percentage increase in sales from 2022 to 2023?" Now you know: you need sales figures for two specific years. Go find exactly those numbers.
Step 2: Identify the data points. Locate the exact numbers you need. For bar charts, trace from the top of the bar horizontally to the y-axis. For tables, find the intersection of the right row and column. Write them down (or remember them).
Step 3: Calculate. Apply the appropriate formula. Percentage increase = (new - old) / old × 100. Don't overcomplicate it.
The harder questions combine multiple operations. Example: "If Company A's profit margin is 15% and their revenue in Q3 was shown in the chart, what was their profit in Q3?" This requires: reading the revenue from the chart, then calculating 15% of that figure.
For multi-step questions, write down each intermediate result. Under time pressure, it's easy to lose track of numbers in your head.
Scale manipulation: A bar chart's y-axis starts at 90 instead of 0. The visual difference between bars looks huge, but the actual difference is tiny. Always read the axis.
Mixed units: The table header says "Revenue (£ millions)" but the question asks for the answer in thousands. You need to convert.
Missing data: The chart shows data for 5 regions but the question asks about "all regions." Is there a sixth region not shown? Check for footnotes or totals.
The best practice is real financial data. Open a newspaper's business section, look at a chart, and ask yourself questions about it. Time yourself. SHL's practice tests provide excellent format-specific practice. For Indian candidates preparing for Cognizant, TCS, or Wipro assessments, AMCAT's practice portal offers relevant question formats.
These examples show the 5-step process applied under a 60-second time budget. The mistakes called out in each are the most common ones graders see in real test responses.
Chart: "Quarterly revenue by region, 2024 (£ millions). North: Q1=12.0, Q2=14.4, Q3=13.2, Q4=15.0. South: Q1=8.0, Q2=8.4, Q3=9.6, Q4=10.5. East: Q1=5.0, Q2=5.5, Q3=6.0, Q4=6.5."
Question: "By what percentage did the South region's Q4 revenue exceed its Q1 revenue?"
Approach: Step 1 — read the legend (regions colour-coded). Step 2 — locate South Q4 (10.5) and South Q1 (8.0). Step 3 — formula: percentage change = (10.5 − 8.0) / 8.0 = 2.5 / 8.0 = 0.3125 → 31.25%. Step 4 — sanity check: 8.0 + 31.25% ≈ 10.5 ✓. Answer: 31.3% (rounded to 1 dp).
The most common error: reading the Q1 value off the wrong region. Bar charts cluster regions side-by-side per quarter; under time pressure candidates grab the leftmost bar in Q4 instead of the South bar. Always anchor on the legend colour first, then trace the column.
Chart: "2024 advertising spend by channel: Search 38%, Social 24%, Display 18%, Video 12%, Other 8%. Total spend £4.0 million."
Question: "If Social and Video spend together is reallocated equally to Search and Display, what is the new Search budget in £ millions?"
Approach: Step 1 — original Search = 38% × £4.0m = £1.52m. Step 2 — Social + Video combined = (24% + 12%) × £4.0m = 36% × £4.0m = £1.44m. Step 3 — half of £1.44m goes to Search: £0.72m. Step 4 — new Search budget = £1.52m + £0.72m = £2.24m. Answer: £2.24 million.
The most common error: forgetting that the question says "equally" between Search AND Display, and giving Search the full reallocation (£2.96m). Read the redistribution instruction at least twice — pie-chart reallocation questions are designed around exactly this kind of misread.
Chart: "Monthly active users (millions) for two competing apps over 12 months. App Alpha: starts at 4.0, rises to 6.2 by month 6, dips to 5.8 by month 9, ends at 6.6 by month 12. App Beta: starts at 3.0, climbs steadily to 6.0 by month 12."
Question: "In which month did App Alpha's monthly user growth rate first turn negative?"
Approach: Step 1 — note that "growth rate turning negative" means month-on-month users decreasing, not the cumulative total. Step 2 — scan the line for the first downward segment. Step 3 — App Alpha rises continuously to month 6 (6.2), so growth is positive through then. Step 4 — between month 6 and month 9, users dropped from 6.2 to 5.8 — so growth turned negative somewhere in this range. Without finer-grained data points the answer is "between month 6 and month 9," typically expressed as "month 7" if the question forces a single month. Answer: month 7.
The most common error: answering "month 9" because that's where the chart bottoms out. The question asks when growth first turned negative — that's the start of the downward slope, not the bottom of it.
Table: "Inventory turnover ratio by store category, 2024:"
| Store category | Avg inventory £k | Annual sales £k | Inventory turnover |
|---|---|---|---|
| Flagship | 320 | 4,800 | ? |
| High street | 180 | 1,800 | 10.0 |
| Concession | 90 | 1,440 | 16.0 |
| Online warehouse | 520 | 10,400 | 20.0 |
Question: "Calculate the Flagship category's inventory turnover. By how much does it differ from the average turnover of the other three categories?"
Approach: Step 1 — definition (visible in row 2): turnover = sales / inventory. Step 2 — Flagship turnover = 4,800 / 320 = 15.0. Step 3 — average of the other three = (10.0 + 16.0 + 20.0) / 3 = 46.0 / 3 = 15.33. Step 4 — difference = 15.33 − 15.00 = 0.33 (or "Flagship is 0.33 below the average"). Answer: Flagship's turnover (15.0) is 0.33 below the average of the other three (15.33).
The most common error: computing the average of all four categories including Flagship, instead of the other three. The phrase "the other three" is the discriminator. Read modifiers like "other," "remaining," "excluding" carefully — they often invert the calculation.
Under timed conditions, working memory contracts. Most candidates' chart-reading accuracy drops 15-25% between a 90-second-per-question pace and a 45-second pace, even on the same questions. The mechanism is well-studied: time pressure narrows attention, which causes you to anchor on the first plausible number you see instead of cross-checking.
Three practical countermeasures:
Pre-read the chart before reading the question. Five seconds glancing at the chart's title, axes, and legend — before you know what's being asked — primes your visual system. When the question appears, your eyes already know where to look.
Write the formula before touching the chart. If the question is "what percentage of the total," write "X / total × 100" on your scratch paper before locating X. This stops you from grabbing the wrong base.
Verify the unit once at the end. Quickly check: did the question want millions or thousands? Percentages or absolute values? The single 3-second check at the end catches the majority of unit-error losses.