Cumulative Frequency Graphs: Where the Easy Marks Hide

If you have ever stared at a cumulative frequency table and wondered "do I plot the point at the midpoint or the end of each group?" — you are asking exactly the right question. It is the single most common mistake at GCSE, and getting it wrong throws every later mark off course. The answer, which we will prove in a moment, is that you plot at the upper class boundary, never the midpoint.

Cumulative frequency is one of the most reliable topics on the Higher paper. The method barely changes from question to question: build a running total, plot it correctly, draw a smooth curve, then read off the median and quartiles. Master those four steps and you have a near-guaranteed handful of marks every series.

This guide walks through what cumulative frequency means, exactly where to plot your points, how to draw the graph step by step, and how to find the median, quartiles and interquartile range — with the mistakes examiners flag most.

What Cumulative Frequency Actually Means

Cumulative frequency is just a running total of the frequencies as you move up through the data. Each value answers one question: how many data items are less than or equal to the upper end of this group?

Take this table of times, in seconds, that 40 students took to complete a puzzle:

To find the cumulative frequency, add each frequency to the total so far:

$$4,\quad 4+9=13,\quad 13+13=26,\quad 26+8=34,\quad 34+6=40$$

So 13 students took 20 seconds or less, 26 took 30 seconds or less, and so on. Notice the final cumulative frequency is 40 — the total number of students. If your last value does not match the total, you have made an arithmetic slip. Checking this one number catches most errors before they cost you marks.

Do You Plot at the Midpoint? The Question That Costs Marks

Here is the rule, and it is worth committing to memory: on a cumulative frequency graph you plot each point at the upper class boundary, not the midpoint.

Why? Because cumulative frequency tells you how many values fall below a certain point. By the time you reach $t = 20$, you know that 13 students have finished — but you only know that at the end of the interval, not halfway through it. The running total is only complete at the top of each class, so that is where the point belongs.

This is different from finding an estimate of the mean, where you do use midpoints. Mixing the two up is the classic trap. So:

• Cumulative frequency graph → plot at the upper boundary (10, 20, 30, 40, 50).
• Estimating the mean → use the midpoint (5, 15, 25, 35, 45).

For our table, that means plotting the points $(10, 4)$, $(20, 13)$, $(30, 26)$, $(40, 34)$ and $(50, 40)$. You should also start the curve at $(0, 0)$, since zero students had finished before any time had passed.

How to Draw a Cumulative Frequency Graph

Follow the same five steps every time:

1. Add a cumulative frequency column to the table (the running total).
2. Plot each point at the upper class boundary against its cumulative frequency.
3. Include the starting point at the lowest boundary with a cumulative frequency of zero — here, $(0, 0)$.
4. Join the points with a smooth curve, not straight ruled lines. Examiners expect a gentle S-shape, not a connect-the-dots zig-zag.
5. Label the axes — time on the horizontal axis, "Cumulative frequency" on the vertical axis.

Using the points above, your curve rises steeply through the middle (where most students finished, between 20 and 30 seconds) and flattens at the top once everyone is accounted for. That characteristic stretched-S shape is your visual check that the plotting is correct.

Finding the Median, Quartiles and Interquartile Range

Once the curve is drawn, the cumulative frequency graph does the hard work for you. With $n = 40$ data values, use these positions on the vertical axis:

$$\text{Median: } \frac{n}{2} = \frac{40}{2} = 20$$ $$\text{Lower quartile: } \frac{n}{4} = \frac{40}{4} = 10$$ $$\text{Upper quartile: } \frac{3n}{4} = \frac{3 \times 40}{4} = 30$$

For each one, find that value on the cumulative frequency axis, draw a horizontal line across to the curve, then drop straight down to read the time.

Reading from our curve:

• A cumulative frequency of 20 meets the curve at roughly $t = 28$ seconds — the median.
• A cumulative frequency of 10 gives a lower quartile of about $t = 17$ seconds.
• A cumulative frequency of 30 gives an upper quartile of about $t = 35$ seconds.

The interquartile range (IQR) measures the spread of the middle half of the data:

$$\text{IQR} = \text{upper quartile} - \text{lower quartile} = 35 - 17 = 18 \text{ seconds}$$

The IQR is especially useful because it ignores the extreme values at either end, so it is not distorted by one unusually fast or slow student. These same quartiles feed straight into box plots, so the skill pays off twice.

Common Mistakes Examiners Look For

Examiner reports flag the same errors year after year on cumulative frequency questions:

• Plotting at the midpoint instead of the upper class boundary — the number one error, and the reason your graph would be marked wrong from the very first point.
• Forgetting the starting point at $(0, 0)$, which leaves the curve floating and distorts readings near the bottom.
• Drawing straight lines between points rather than a smooth curve.
• Reading the quartile positions wrong — remember the positions $\frac{n}{2}$, $\frac{n}{4}$ and $\frac{3n}{4}$ are read on the vertical (cumulative frequency) axis, then traced down to the horizontal axis.
• Not checking the final total — if your last cumulative value does not equal $n$, something has gone wrong earlier.

Most of these are avoidable with a single habit: pause after plotting and ask, "have I used the end of each group, and does my last point reach the total?"

💡 Practise now: Cumulative frequency questions on Bow Tie Maths — the app builds a Topic Radar from your answers so you always know where to focus next.

Summary

Cumulative frequency is a running total, and the golden rule is to plot every point at the upper class boundary, never the midpoint — midpoints belong to estimating the mean, not to this graph. Draw a smooth curve from $(0, 0)$, then read the median, quartiles and interquartile range from the positions $\frac{n}{2}$, $\frac{n}{4}$ and $\frac{3n}{4}$ on the vertical axis. Always finish by checking your last cumulative value equals the total.

If you want to put this into practice, try Bow Tie Maths — it generates questions on this topic at your level and tracks your progress over time.