Data interpretation mastery: tables, graphs, and charts
Data interpretation is the single most common question type in numerical reasoning tests. You are given a table, bar chart, line graph, or pie chart and asked to calculate, compare, or extrapolate. The data is always sufficient — the challenge is finding it fast. If you need the wider method first, start with numerical reasoning tests; if the arithmetic is the bottleneck, use percentage shortcuts.
- Read the question before the chart. Otherwise you waste time scanning irrelevant data.
- Units and bases decide most answers: thousands vs millions, monthly vs annual, old value vs new value.
- Tables punish row/column mistakes; charts punish visual assumptions.
- The best review habit is tagging the miss: wrong unit, wrong base, wrong row or rushed arithmetic.
Data interpretation speed comes from timed chart drills, not rereading methods. Start numerical practice
The reading order that saves time
- Read the question first. Know what you need before looking at the data.
- Read titles, axis labels, and units. Is it thousands or millions? Quarterly or annual?
- Locate the relevant data points. Ignore everything else.
- Calculate and sanity-check. Does the answer feel right given the scale?
What each chart type is really testing
| Format | What it tests | Main trap |
|---|---|---|
| Table | Finding the right row, column and unit quickly. | Reading the adjacent row or mixing periods. |
| Bar chart | Comparing categories and calculating shares. | Using the largest bar as the denominator instead of the total. |
| Line graph | Trend, change and rate over time. | Missing that the axis does not start at zero. |
| Pie chart | Turning percentages into underlying values. | Forgetting that all slices share the same total. |
Regional revenue ($ millions)
Total = 42 + 38 + 56 + 14 = 150. APAC share = 42 / 150 = 28%.
Trap: candidates often calculate 42 / 56 (comparing to the largest region) instead of 42 / 150 (share of total).
Table questions
Tables are the densest format. They pack a lot of data into a small space, which means the main risk is reading the wrong row or column. Use your finger (or cursor) to track across rows.
A good table routine is mechanical: find the row label, find the column label, check the unit, then calculate. Do not start calculating while your eyes are still searching. That is how candidates copy the 2024 figure when the question asks for 2025, or use revenue when the question asks for profit.
Growth = (4.0 − 3.2) / 3.2 = 0.8 / 3.2 = 25%.
Common error: dividing by the new value (0.8 / 4.0 = 20%). Always divide by the base (starting) value.
Chart-specific traps
- Bar charts: stacked vs grouped. Stacked bars require subtraction to get individual values.
- Line graphs: the y-axis may not start at zero, exaggerating trends.
- Pie charts: you often need to compute the underlying value from a percentage and a given total.
- Dual-axis charts: the left and right axes can use different units. Check which series belongs to which axis.
- Index charts: an index of 120 means 20% above the base, not a raw value of 120.
Worked example: index values
Suppose a sales index is set to 100 in 2024 and rises to 115 in 2025. That is a 15% increase from the base year. If 2024 sales were $80m, then 2025 sales are 80 x 1.15 = $92m. The index is not dollars, units or customers. It is a relative scale.
How to review data interpretation misses
Split every wrong answer into one of four buckets: wrong data point, wrong denominator, wrong unit, or arithmetic slip. If your misses are mostly wrong data points, slow your first five seconds down. If they are mostly arithmetic slips, keep the method but add estimation before choosing an option. For a mixed drill after this page, use the 25 worked numerical questions.
Source notes
forge checks data-interpretation guidance against official SHL numerical material, including SHL’s numerical reasoning example questions. Those examples reinforce the same point: facts and figures in tables are the core input, and the main skill is using the right data under time pressure.
Drill data interpretation on forge
323 data interpretation questions across SHL, AON, Cappfinity, and forge — each tagged to a specific sub-skill.
Start practisingFrequently asked questions
What is data interpretation in numerical reasoning?+
Data interpretation questions ask you to read tables, charts or graphs, locate the relevant figures, and calculate a comparison, percentage, rate or trend.
What is the biggest data interpretation trap?+
The biggest trap is using the wrong row, unit, time period or denominator. Most mistakes happen before the arithmetic starts.
How do I get faster at chart questions?+
Read the question first, check units and axes, locate only the relevant data, then calculate. Speed improves when you stop scanning the whole chart.
Stop reading. Start practising.
Try a few real questions — free.
You know the format now. Reading about the test won't move your score — timed reps will. Do a free diagnostic, earn credits just for signing up, and see exactly where you stand.