Questions built around a chart or a table reward a skill that is easy to underrate: reading quantitative information carefully instead of glancing at it. Under time pressure it is tempting to see a rising line or a big number and answer from that first impression — which is exactly how a well-made question catches you. A graph is not a picture to admire; it is a small dataset you have to query. This article sets out a steady, repeatable way to work through the graphs, tables and diagrams you may meet in the general academic material so you spend your time reading the data rather than second-guessing it.
Read the frame before the data
Your first job with any chart is not to read the values — it is to understand what the chart is measuring. Read the title, then both axis labels, then the units, then any legend. A surprising number of avoidable errors come from confusing thousands with millions, percentages with counts, or one category's line with another's. Thirty seconds spent orienting yourself to the frame protects every answer you draw from it.
Pay particular attention to units and scale. An axis that starts at a value other than zero can make a small change look dramatic; a logarithmic scale compresses large differences. The chart is not trying to trick you, but a careless reader tricks themselves. Note whether numbers are absolute or relative, and whether a table's rows are the thing being compared or the thing being held constant. This is the same "map before detail" habit that works for text, described in our reading comprehension strategy: understand the structure first, then look up specifics with confidence.
Let the question tell you what to extract
Once you understand the frame, turn to the question and decide precisely what it asks before you go hunting through the data. Does it want a single value, a difference between two points, a trend over time, or a proportion of a whole? Naming the task keeps you from reading the entire table when you only need one cell, and from computing an average when the question asked for a maximum.
Then return to the graph or table and extract only what you need. For a specific value, trace carefully from the axis to the data point and back — do not estimate when the figure is printed. For a comparison, locate both quantities before you judge which is larger, because eyeballing a bar chart is where confident mistakes live. For a trend, describe the shape in plain words ("rising, then flat after the third year") before you pick an option. Answering from what the data actually shows, not from a plausible story about it, is the core discipline here — closely related to avoiding the small slips covered in how to avoid careless mistakes.
Handle the common question shapes deliberately
A few shapes recur, and each rewards a specific move:
- Read-off — find one value at one point; trace both axes rather than guessing from position.
- Difference or change — read both values exactly, then subtract; do not judge the gap by eye.
- Trend — summarise the direction and any turning points in a phrase before comparing options.
- Proportion or share — check whether the figure is already a percentage or whether you must divide by a total.
- Combine sources — when a question spans a chart and a table, confirm both use the same units and categories before you merge them.
Predicting the answer in your own terms before you read the options is as protective here as it is anywhere: it stops a distractor that misuses the same numbers from steering you. When the arithmetic gets fiddly, a quick estimate often rules out two or three options at once, which is exactly the move described in our guide to using estimation on the dMAT.
Manage time and rough work
Data questions can quietly drain your clock if you re-read the whole chart for every part. Orient yourself to the frame once, then treat each sub-question as a targeted lookup. Jot the one or two figures you have extracted on your rough paper so you are not re-tracing the same line, and label them so you do not confuse two similar values later. If a single item resists, mark it, take the best-supported option and move on — pacing protects your total, as with any subtest; see our note on time management under pressure.
You build this skill the same way you build the others: short, deliberate practice. Take one chart, write down what each axis and unit means, then answer a handful of questions by pointing to the exact values you used and checking that they really say what you claimed. Over time you will read data faster not by rushing but by knowing what to look at first.
Keep the purpose in view
Interpreting graphs and tables is not about admiring the picture — it is about querying a small dataset with a clear question in mind: read the frame, let the question tell you what to extract, and answer from the numbers on the page rather than your first impression of them. When you would like to rehearse under realistic conditions, work through original questions in the free dMAT diagnostic, and confirm which task types you are likely to meet on the syllabus.
