WhatsApp us
IGCSE Maths18 Nov 2026

IGCSE Scatter Diagrams and Correlation: Plotting, Describing and Using a Line of Best Fit (0580 E9.5)

Scatter diagrams for Cambridge IGCSE 0580 (E9.5, Core and Extended): how to plot points, describe positive, negative and zero correlation, draw a ruled line of best fit by eye using the mean point, make estimates and explain when they are unreliable.

IGCSE Scatter Diagrams and Correlation: Plotting, Describing and Using a Line of Best Fit (0580 E9.5)

A scatter diagram plots pairs of values, such as hours of revision and test score, as crosses on a grid, so you can see whether the two measurements move together. In Cambridge IGCSE Mathematics 0580 this is sub-topic E9.5, and it is the same at Core and Extended: plot the points, describe the correlation as positive, negative or zero, then draw a straight line of best fit by eye and use it to estimate.

It is one of the most reliable sources of marks in the Statistics strand, provided you follow the drawing rules in the syllabus notes. Those rules are more precise than most students realise, and this guide works through them with four original examples. For the rest of the strand, see the map of all 0580 statistics topics.

What E9.5 says

The syllabus lists three skills:

  1. Draw and interpret scatter diagrams.
  2. Understand positive, negative and zero correlation.
  3. Draw by eye, interpret and use a straight line of best fit.

Its notes then say how the line should look. It should be a single ruled line drawn by inspection, it should extend across the full data set, and it does not have to pass through any particular point, but there should be a roughly even spread of points either side along its whole length. Points should be clearly marked, for example as small crosses.

There is no calculation of a correlation coefficient or regression equation in 0580. If you want those, they belong to 0607 and to later courses. Students who would like this checked against their own working can book Cambridge IGCSE maths lessons built around exam-style questions.

Correlation in words

Pattern on the diagramNameExample in words
Points rise from left to rightPositive correlationMore revision, higher scores
Points fall from left to rightNegative correlationWarmer days, lower heating bills
No clear directionZero (no) correlationShoe size and maths score
Points close to a lineStrongMost points sit near the line
Points loosely around a lineWeakA trend, with a lot of scatter

An exam answer should name the type of correlation, and, if asked to interpret, put it in context: "As revision time increases, test score tends to increase." The word "tends" matters, because correlation describes a trend, not a rule for every individual.

Worked example 1: plotting and describing

Ten students recorded revision time and test score.

Revision (hours)1223455678
Score35404650555864667278

Plotting. Put the variable you control or measure first (revision) on the horizontal axis and the score on the vertical axis. Choose scales that use most of the grid: 0 to 8 across, 30 to 80 up. Mark each point with a small cross. Two students both revised for 2 hours, so there are two crosses in that column, at 40 and 46.

Describing. The crosses rise from left to right and lie close to a straight line, so there is strong positive correlation. In context: students who revised for longer tended to score higher.

Worked example 2: drawing a line of best fit and estimating

Draw a line of best fit for the data above and estimate the score of a student who revised for 4.5 hours.

Step 1: find the mean point. Mean revision = 43 ÷ 10 = 4.3 hours. Mean score = 564 ÷ 10 = 56.4. Plot (4.3, 56.4) with a different mark, such as a small circle.

The syllabus does not require the line to pass through the mean point, but it is a strong guide: a straight line through it, angled to follow the trend, almost always gives the even spread of points either side that examiners look for.

Step 2: rule the line. Lay the ruler through the mean point and rotate it until roughly half the crosses are above and half below along the whole length. One sensible line passes close to (1, 37) and (8, 79). Draw it across the full data set, from the first point to the last, as a single ruled line.

Step 3: estimate. Go up from 4.5 on the horizontal axis to the line and across: about 58 marks.

Because each student draws their own line by eye, a small range of estimates is normal. What earns the mark is reading accurately from your own line, with the guide lines shown.

Worked example 3: when an estimate is unreliable

Use the line to estimate the score of a student who revised for 12 hours. Comment on your answer.

Extending the line (it rises by about 6 marks per hour) gives roughly 31 + 6 × 12, which is about 103. The test is marked out of 100, so this is impossible.

The data only covers 1 to 8 hours. Estimating inside that range (interpolation) is reasonable. Estimating outside it (extrapolation) assumes the trend continues, and here it clearly cannot. A full-mark comment says both things: the value is outside the range of the data, so the estimate is unreliable, and the context shows why.

Worked example 4: an outlier and a causation trap

A survey of eight homes gives the average outdoor temperature (°C) and the monthly heating cost (in some currency).

Temperature (°C)24578101214
Heating cost9084807240645850

(a) Describe the correlation. (b) Identify the point that does not fit. (c) A student says, "Higher temperatures cause lower costs, so this data proves it." Comment.

(a) As temperature rises, cost falls: negative correlation, fairly strong apart from one point.

(b) (8, 40) sits well below the pattern; the trend suggests a cost near 70 at 8 °C. Perhaps that household was away. When drawing a line of best fit, ignore an outlier like this rather than letting it pull the line down.

(c) Correlation alone does not prove that one variable causes the other. Here a link is plausible, but the data only shows that the two tend to move together, from just eight homes. This is the E9.2 skill of recognising "restrictions on drawing conclusions".

For contrast, a scatter diagram of shoe size against maths score for a class would show crosses with no direction at all: zero correlation, and no line of best fit would be useful.

The drawing checklist

  • ☐ Small, clearly marked crosses (the example the syllabus gives), not blobs
  • ☐ Both axes labelled, with even scales
  • ☐ One straight line, ruled, not a curve and not a dot-to-dot zigzag
  • ☐ Line runs across the full set of points, not just the middle
  • ☐ Roughly even spread of points either side along the whole line
  • ☐ Estimates read from your line, with dashed guide lines shown

Calculator and non-calculator papers

Scatter diagram questions need almost no arithmetic, so they appear on both papers. On Paper 1 or Paper 2, the mean point may be the only calculation, and it is chosen to divide neatly. On Paper 3 or Paper 4 you may use the calculator for the mean point, but plotting and drawing are done by hand either way. A calculator with a statistics mode is not needed for 0580.

Mistakes our tutors see

  1. Joining the points with straight segments, like a line graph.
  2. Forcing the line through the origin when the data does not start there.
  3. A line that stops in the middle of the data, or starts at the first axis.
  4. Describing correlation as "positive" without saying what increases with what.
  5. Treating an extrapolated value as reliable.
  6. Saying "correlation proves cause".
  7. Drawing two lines, or a line in pen over a pencil line that was rubbed out badly.

When to get help

  • Your lines of best fit are marked wrong and you cannot see why.
  • You confuse a scatter diagram with a line graph or a bar chart.
  • "Comment on the reliability" questions keep scoring zero.
  • You can describe correlation but cannot read an estimate from your own line.

Self-check

  1. As a car gets older, its value goes down. What type of correlation is this?
  2. A line of best fit passes through (2, 10) and (10, 50). Estimate y when x = 6.
  3. Why should you not join the points on a scatter diagram?
  4. Find the mean point of (2, 5), (4, 9), (6, 10) and (8, 16).
  5. Give an example of two variables with zero correlation.

Answers. 1: negative correlation. 2: the gradient is 40 ÷ 8 = 5, so y = 10 + 5 × 4 = 30. 3: the points are separate individuals; the trend is shown by one straight line of best fit, not by joining them. 4: (20 ÷ 4, 40 ÷ 4) = (5, 10). 5: for example, house number and height of the person living there.

Where it leads

0607 International Mathematics asks for a linear regression equation on the graphic display calculator at Extended level, and our guide to regression on the GDC for 0607 shows how that compares with drawing by eye. In the IB Diploma, bivariate data sits inside "Statistics and probability". Within 0580, the natural next topics are the cumulative frequency curve and quartiles and averages and spread for grouped data. Our pie, bar and stem-and-leaf guide covers the other diagrams in the strand.

How Maths Bodhi helps

Scatter questions are lost at the ruler, not the calculator. Our tutors have the student draw a line of best fit live, then talk through why it sits where it does, so the habits stick before mocks. Lessons use original exam-style data sets with a written record of each attempt. Read how our IGCSE maths tutors plan sessions.

Home tuition is available in Gurgaon (Gurugram). Students elsewhere in India and abroad learn with us online. You can book a free demo class.

Written and checked by the Maths Bodhi team. Facts checked against official Cambridge International and Pearson sources, November 2026.

Frequently asked questions

Are scatter diagrams in IGCSE Core and Extended?
Yes. In 0580, scatter diagrams (E9.5) are the same at both tiers: plot points, describe correlation and draw and use a straight line of best fit by eye.
Does the line of best fit have to go through the origin?
No. It should be a single ruled line across the full data set with points roughly evenly spread either side. It only passes through the origin if the data suggests it.
Does the line of best fit have to go through the mean point?
The syllabus does not require it, but drawing the line through the mean point is a reliable way to get an even spread of points either side.
What are the three types of correlation in IGCSE maths?
Positive (both increase together), negative (one increases as the other decreases) and zero (no clear relationship). You may also describe it as strong or weak.
Why is extrapolation unreliable on a scatter diagram?
It assumes the trend continues beyond the data you have. Outside the range of the data, the relationship may change, as in a score that would go above the maximum mark.
Do I need to calculate a correlation coefficient in IGCSE 0580?
No. 0580 asks you to describe correlation in words and draw a line of best fit by eye. Regression on a graphic calculator is part of 0607 Extended.
IGCSE scatter diagramscorrelation IGCSEline of best fit by eye0580 E9.5IGCSE maths tutor

Related guides

Keep exploring this maths topic

IGCSE Maths18 Nov 2026

IGCSE Maths Statistics Topics (0580 E9): All Seven Sub-topics, Core vs Extended, and How They Are Examined

A map of the Statistics strand in Cambridge IGCSE Mathematics 0580: the seven E9 sub-topics with their Core and Extended status, what each one asks you to do, how statistics appears on the non-calculator and calculator papers, and a study order.

IGCSE statistics0580 E9 statisticscumulative frequency and histogramsCore vs Extended statistics

By Maths Bodhi Team

Read guide

IGCSE Maths18 Nov 2026

0607 Regression on the GDC: Data Lists, Averages and the Line y = ax + b

A hands-on guide to the statistics functions a 0607 student must use on a graphic display calculator: entering lists, mean, median and quartiles, grouped means and, at Extended, the linear regression equation, with four worked examples and the interpretation marks students miss.

0607 regression GDClinear regression IGCSE0607 averages on a calculatorIGCSE International Maths statistics

By Maths Bodhi Team

Read guide

IGCSE Maths18 Nov 2026

IGCSE Statistics Questions: 11 Practice Questions on Averages, Grouped Data, Cumulative Frequency and Histograms

Eleven original IGCSE 0580 statistics questions in two parts: averages, spread and grouped data, then pie charts, cumulative frequency, histograms, scatter diagrams and comparing distributions, each with working and credit notes.

IGCSE statistics practiceIGCSE histogram questionsIGCSE cumulative frequency questions0580 statistics

By Maths Bodhi Team

Read guide

IGCSE Maths18 Nov 2026

IGCSE Cumulative Frequency: Tables, Curves, Median, Quartiles, IQR and Percentiles (0580 E9.6)

Cumulative frequency for Cambridge IGCSE 0580 Extended (E9.6): building the running-total table, plotting at upper class boundaries, drawing the smooth curve, and reading the median, quartiles, interquartile range and percentiles, with four worked examples and a self-check.

IGCSE cumulative frequencyquartiles from a graphinterquartile range IGCSE0580 E9.6 Extended

By Maths Bodhi Team

Read guide

IGCSE Maths18 Nov 2026

IGCSE Maths Command Words: What "Show That", "Write Down" and "Sketch" Really Ask

The command words used in Cambridge IGCSE 0580 maths papers, explained in plain English, with how much working each one needs, original worked examples, a practice plan and a checklist.

IGCSE maths command words0580 show that questionsIGCSE write down meaningIGCSE exam technique

By Maths Bodhi Team

Read guide

IGCSE Maths18 Nov 2026

IGCSE Statistics Charts: Pie Charts, Bar Charts, Pictograms and Stem-and-Leaf Diagrams (0580 E9.4)

Every chart in Cambridge IGCSE 0580 topic E9.4, the same list for Core and Extended: pie charts, composite and dual bar charts, pictograms, stem-and-leaf diagrams and frequency distributions, with four worked examples, the drawing rules examiners look for, and a self-check.

IGCSE pie chartsstem-and-leaf diagramscomposite bar charts0580 E9.4

By Maths Bodhi Team

Read guide