Why a histogram uses area, not height
A bar chart shows categories — favourite colours, pets, makes of car. The bars have gaps between them and equal widths, and the height of each bar is the frequency.
A histogram shows continuous data that has been grouped into classes — times, masses, heights, distances. The classes sit next to each other on a number line, so the bars touch: no gaps. And here is the twist that Paper 4 tests every year: the classes are often different widths. One class might cover 5 seconds and its neighbour 20 seconds.
Suppose 10 runners finish in the class 25 < t ≤ 45 and 30 runners finish in 10 < t ≤ 15. If you plotted raw frequency as height, the 20-minute-wide class would tower over the graph with an enormous area for only 10 runners, while 30 runners get squeezed into a thin sliver. The picture would lie about where the data is crowded.
Height = frequency ✗ misleading
The wide class grabs a huge area for just 10 values. The eye reads area, so the graph misleads.
Height = frequency density ✓ honest
Same data, height scaled by class width. Now every bar’s area equals its frequency, and the crowding is where it really is.
Frequency density
Class width means upper class boundary − lower class boundary. For the class 10 < t ≤ 15 the width is 15 − 10 = 5. Units help you think: if t is in minutes, a frequency density of 6 means “6 people per minute of the scale” — a rate, exactly like speed or price per kg.
Example 1The masses, m grams, of 60 apples are grouped as 60 < m ≤ 100 (16 apples), 100 < m ≤ 120 (24), 120 < m ≤ 150 (15), 150 < m ≤ 200 (5). Work out the frequency density of each class.
fd = 24 ÷ 20 = 1.2
fd = 15 ÷ 30 = 0.5
fd = 5 ÷ 50 = 0.1
| Mass m (g) | Frequency | Class width | Frequency density |
|---|---|---|---|
| 60 < m ≤ 100 | 16 | 40 | 16 ÷ 40 = 0.4 |
| 100 < m ≤ 120 | 24 | 20 | 24 ÷ 20 = 1.2 |
| 120 < m ≤ 150 | 15 | 30 | 15 ÷ 30 = 0.5 |
| 150 < m ≤ 200 | 5 | 50 | 5 ÷ 50 = 0.1 |
Drawing a histogram
The routine never changes. For each class: width, then frequency density, then draw a bar from lower bound to upper bound at that height. The bars touch, and the horizontal axis is a proper continuous scale (equal gaps for equal amounts), not a list of labels.
- Add a class width column and a frequency density column to the table.
- Choose a vertical scale that fits the largest frequency density comfortably.
- Draw each bar exactly between its class boundaries — no gaps, no overlaps.
- Label the axes: the variable (with units) across, Frequency density up.
Times taken to solve a puzzle by 85 students — the finished histogram
Data: 0 < t ≤ 10 f = 20, 10 < t ≤ 15 f = 30, 15 < t ≤ 25 f = 25, 25 < t ≤ 45 f = 10. Check any bar: area = width × fd gives back the frequency.
Histogram builder
Edit the class boundaries and frequencies — the frequency densities and the histogram redraw instantly. Try making one class very wide, or create an overlap and see it flagged.
Reading a histogram back into a table
Examiners love the reverse question: you are given the drawn histogram and asked to recover the frequencies. Rearrange the definition:
Read the height of each bar from the vertical scale (read it carefully — count the small squares per unit), multiply by the width of the class, and you have the frequency. Because frequency is an area, you can even take part of a bar.
Part of a bar: estimating within a class
If a bar spans 10 < x ≤ 30, the data inside it is assumed to be spread evenly. So the number of values between 15 and 25 is estimated by the area of just that slice: fd × (25 − 15). This is an estimate — say so in your answer — because we cannot know exactly where each value sits inside its class.
Example 2A histogram of waiting times has bars: 0 < t ≤ 20 at height 0.9, 20 < t ≤ 30 at height 3.2, 30 < t ≤ 35 at height 6, 35 < t ≤ 60 at height 0.8. Complete the frequency table and find the total number of people.
20 < t ≤ 30: 3.2 × 10 = 32
30 < t ≤ 35: 6 × 5 = 30
35 < t ≤ 60: 0.8 × 25 = 20
More worked examples
Example 3On a histogram of parcel masses, the bar for 20 < m ≤ 60 has frequency density 0.35. (a) How many parcels are in this class? (b) Estimate how many parcels have a mass between 30 kg and 50 kg.
Example 4A histogram is drawn on paper with no vertical scale. The class 0 < x ≤ 5 has frequency 15 and its bar is 6 cm tall. The bar for 5 < x ≤ 15 is 2.4 cm tall. Find the frequency of the second class.
Example 5Journey times are grouped as 0 < t ≤ 30 (30 journeys), 30 < t ≤ 40 (25) and 40 < t ≤ 60 (28). A student says the modal class is 0 < t ≤ 30 because 30 is the biggest frequency. Show that the student is wrong.
fd = 25 ÷ 10 = 2.5
fd = 28 ÷ 20 = 1.4
Your turn
Practice set
- Find the frequency density of the class 10 < x ≤ 25 with frequency 45.
- A histogram bar for 20 < x ≤ 50 has height 1.8. Find the frequency.
- Times: 0 < t ≤ 4 f = 10, 4 < t ≤ 10 f = 18, 10 < t ≤ 20 f = 15. Work out all three frequency densities and state the modal class.
- The bar for 30 < x ≤ 70 has frequency density 0.6. Estimate the number of values between 40 and 55.
- A survey of 50 people gives classes 0 < x ≤ 10 (fd = 1.2), 10 < x ≤ 15 (fd = 3.6) and 15 < x ≤ 35 with an unknown frequency. Find the height of the third bar.
- Width = 15, so fd = 45 ÷ 15 = 3
- f = 1.8 × 30 = 54
- 10 ÷ 4 = 2.5, 18 ÷ 6 = 3, 15 ÷ 10 = 1.5; modal class 4 < t ≤ 10
- Slice width 55 − 40 = 15, so 0.6 × 15 = 9 (an estimate)
- Known frequencies: 1.2 × 10 = 12 and 3.6 × 5 = 18; missing f = 50 − 12 − 18 = 20 over width 20, so height = 20 ÷ 20 = fd = 1
Key points to remember
- A histogram shows continuous grouped data: bars touch, the horizontal axis is a real number line, and area represents frequency.
- frequency density = frequency ÷ class width — and the reverse, frequency = fd × width, reads the histogram back.
- The vertical axis must be labelled Frequency density. Plotting raw frequency as height is the number-one error on this topic.
- Class width = upper boundary − lower boundary. Work out a width column first, every time.
- Part of a bar gives an estimate of the frequency in part of a class: fd × slice width, assuming even spread.
- The modal class is the class with the highest frequency density — the tallest bar — not the biggest frequency.
- Missing-bar questions: find every known frequency, subtract from the total, then divide by the missing width to get the height.
Where next?
🎥 Take the interactive module
Representing Data — the full teaching module for this topic, with real past-paper questions and mark schemes.
✏️ Practise exam questions
Statistics — 25 exam-style questions worth 99 marks, with full answers and the mark scheme.
Progress is saved on this device only and appears on the notes home page.