Imagine a page full of numbers: the rainfall for every day of a week, a shop's sales for six months, or the answers of 200 people to one survey question. Reading a long list of numbers tells you the facts, but it hides the pattern. A graph turns those numbers into a picture, so the pattern jumps out at a glance: which day was wettest, whether sales are rising, or which answer was the most popular. This guide explains the three graphs you meet most often — bar charts, line charts and pie charts — how to read each one, and the mistakes that catch people out.
What a graph is and why we use one
A graph is a picture of numbers. Every graph has the same basic parts. There is a title that says what the graph is about, and labels that say what is being measured. Most graphs have two axes: a horizontal axis along the bottom and a vertical axis up the side. The scale on an axis is the size of each step, for example steps of 5 millimetres or 50 sales. Reading a graph well is mostly about two habits: always read the title and the axis labels first, and always check the scale before you compare anything.
Graphs are useful because our eyes are good at comparing shapes and sizes. A tall bar next to a short bar says "much bigger" instantly, in a way that the numbers 18 and 3 do not. A line climbing steadily says "rising", and a pie slice that takes up a third of the circle says "about a third of the whole". The rest of this guide shows how each type of graph does that job, with one worked example for each.
Bar charts: comparing at a glance
A bar chart compares amounts in different categories. Each category gets one bar, and the length or height of the bar shows the amount. The bars are usually drawn with gaps between them, because the categories are separate things — days of the week, subjects, or types of transport — not a continuous flow. To read a value, find the category along the axis, follow its bar to the top, and read across to the number on the other axis.
The scale matters more than anything else in a bar chart. If each step on the axis stands for 5 millimetres, a bar that reaches the "10" line means 10 millimetres, not 10 bars. When bars are close in height, read carefully against the gridlines rather than guessing by eye. And remember that the bar's length is the whole message: a bar twice as long as another means twice the amount, provided both bars start at zero.
Worked example 1: reading a bar chart of weekly rainfall
A weather station records rainfall, in millimetres, for one week. The bar chart shows one bar for each day:
- Monday: 4 mm
- Tuesday: 12 mm
- Wednesday: 7 mm
- Thursday: 0 mm
- Friday: 18 mm
- Saturday: 9 mm
- Sunday: 3 mm
Three readings come straight from the bars. The wettest day is Friday, at 18 mm, because its bar is the tallest. The driest day is Thursday, at 0 mm, because it has no bar at all — a bar of height zero. The difference between the wettest and driest days is 18 − 0 = 18 mm. Two days had more than 10 mm of rain: Tuesday (12 mm) and Friday (18 mm). Adding the week together gives 4 + 12 + 7 + 0 + 18 + 9 + 3 = 53 mm in total, so the average day saw 53 ÷ 7 = 7.57 mm, about 7.6 mm. Notice how the average (7.6 mm) sits below Friday's 18 mm: one very wet day can pull the total up without making every day wet.
Line charts: change over time
A line chart shows how something changes, usually over time. Points are plotted for each time period and joined with a line, so the eye follows the rise and fall. Unlike a bar chart, the points are joined because the thing being measured flows from one period into the next — time does not stop between March and April.
To read a line chart, pick a point on the line and read its value from the axis, just as with a bar chart. To find a change, subtract the earlier value from the later one: a positive answer means a rise, a negative answer means a fall. The steepness of a section tells you how fast the change happened. A steep climb means a big change in a short time, a gentle slope means a slow change, and a flat section means no change at all over that period — not that the value is zero.
Worked example 2: six months of shop sales
A small shop records how many items it sells each month:
- January: 120
- February: 150
- March: 180
- April: 140
- May: 210
- June: 260
Reading values is simple: sales in March were 180 items. The changes, worked out step by step, tell the story. From January to February, sales rose by 150 − 120 = 30. From February to March, they rose by 180 − 150 = 30 again. From March to April, they fell by 180 − 140 = 40 — the only fall in the six months, and the line slopes downwards there. From April to May, they jumped by 210 − 140 = 70, the steepest climb on the chart. From May to June, they rose by 260 − 210 = 50. Over the whole period, sales grew by 260 − 120 = 140 items. The line chart makes this story obvious in a way a bare list does not: steady growth, one dip in April, then the sharpest rise of all in May.
Pie charts: parts of a whole
A pie chart shows how a whole is shared out. The full circle stands for the whole group — 100% of it — and each slice is one part of that whole. A bigger slice means a bigger share. Pie charts work best with a small number of categories, usually between two and six, because too many thin slices become hard to tell apart.
To turn counts into a pie chart, divide each count by the total and multiply by 100 to get its percentage. If that conversion feels rusty, the guide to fractions, decimals and percentages explains it step by step, and the Percentage Calculator checks your working. The key check is that the percentages must add up to 100%. If they do not, something has gone wrong: a category is missing, a count is wrong, or the total is wrong. That single check catches more pie-chart errors than any other habit.
Worked example 3: a survey of 200 people
A survey asks 200 people how they usually travel to work. The answers are:
- Walk: 60 people
- Cycle: 40 people
- Bus: 70 people
- Car: 30 people
First, check the total: 60 + 40 + 70 + 30 = 200 people, which matches the survey size. Now work out each percentage step by step. Walking is 60 ÷ 200 × 100 = 30%. Cycling is 40 ÷ 200 × 100 = 20%. Travelling by bus is 70 ÷ 200 × 100 = 35%. Travelling by car is 30 ÷ 200 × 100 = 15%. Now the check: 30 + 20 + 35 + 15 = 100%, exactly as it must be. Reading the chart is then easy: the bus is the most common choice at 35%, walking and cycling together make 30 + 20 = 50%, exactly half of all travellers, and car users are the smallest slice at 15%.
Choosing the right chart
Pick the chart that matches the question you are asking. If the question is "which category is biggest, and by how much?", use a bar chart: bars are the easiest shapes to compare side by side. If the question is "how did this change over time?", use a line chart: the joined line shows rises, falls and flat spells. If the question is "how is the whole shared out?", use a pie chart: slices show each part as a share of 100%. A bar chart can also show shares, and a line chart can show categories in order, but each chart is at its clearest when it is used for the job it was made for. If a pie chart would need more than about six slices, a bar chart is usually the clearer choice.
Common mistakes to avoid
1. Reading the wrong axis or ignoring the scale. Always check what each step on the axis stands for. A bar reaching the "10" line means 10 units only if each step is 1 unit; if each step is 5 units, that same position means 50.
2. Being fooled by a truncated axis. If a bar chart's axis starts at 50 instead of 0, a bar for 60 will look far longer than a bar for 55, even though the difference is small. Honest bar charts start at zero, because the bar's length is supposed to show the full amount.
3. Pie slices that do not add to 100%. If your percentages add to 97% or 103%, a slice is missing or a count is wrong. Go back to the counts and the total before trusting the chart.
4. Comparing bars from different scales. Two charts that look alike can have completely different scales. A bar that looks tall on one chart may stand for less than a short bar on another. Compare numbers, not just pictures.
5. Mistaking a flat line for zero. On a line chart, a flat section means the value stayed the same, not that it fell to zero. A line flat at 200 means 200 every month, not nothing at all.
Frequently asked questions
Can a line chart show something other than time? Yes, it can show any ordered sequence, such as scores after each round of a game. But time is its most common use, because the joined line suits things that flow from one period to the next.
Do all pie charts need percentages? No. A pie chart can label its slices with counts instead, for example "60 people". Percentages are popular because they make shares easy to compare, but the slices must still add up to the whole, whether you show counts or percentages.
What if a bar chart's axis does not start at zero? Treat it with care. The bar lengths will exaggerate the differences between categories. If you need an honest comparison, note the axis start and read the numbers rather than judging by eye.
How do I choose a scale for my own graph? Pick a step size that spreads the data across the axis without crowding it. If your largest value is 260, steps of 50 give a readable axis; steps of 10 would be cramped and steps of 200 would hide the differences. Aim for an axis with roughly 5 to 10 steps.
Graphs reward one habit above all: read the labels before you trust the picture. Check the title, check the axes, check the scale — and for a pie chart, check that the parts add up to the whole. With those checks, a bar chart, a line chart or a pie chart will tell you its story honestly, and you will spot the ones that do not.