Use a scatter graph to see how two sets of data change together.
Key words
correlation
how one set of values relates to another
line of best fit
straight line that most points follow
positive correlation
both variables increase together
negative correlation
one variable increases while the other decreases
How to do it
Identify which axis holds each variable.
Look at how points move when one variable changes.
Watch out for
Not drawing any line, or drawing it straight through the origin
Reading off wrong values because scales are not matched
Remember: Correlation does not prove one test causes the other to score higher.
Scatter graphs and correlation Core
In a scatter graph two variables are plotted as points. A line of best fit can be drawn to summarise the trend, which is then used for predictions.
Key words
positive correlation
as one variable increases the other also increases
negative correlation
as one variable increases the other decreases
no correlation
the points show no obvious line
weak correlation
points are loosely spread around a line, giving less precise predictions
How to do it
Write down the two variables, decide which will be horizontal (x) and vertical (y). Choose scales from the minimum to maximum values; use a false origin if the data do not start at zero.
Plot each point accurately on the paper; double‑check that the tick marks are spaced correctly.
Identify the overall trend: look for an upward/rightward slope for a positive correlation, downward/rightward for negative, or no clear line for none.
Decide whether the correlation is weak or strong by judging how close most points lie to a straight line.
If a clear trend exists, draw a line of best fit that roughly follows it; this line need not pass through the origin.
Use the fitted line to interpolate within the data range or extrapolate beyond, but note the risk of error.
Watch out for
Assuming a line of best fit must start at the origin.
Reading a value directly from the grid when the graph is not perfectly aligned; always use the fitted line for interpolation.
Ignoring an outlier and still claiming no correlation; a single point does not negate a clear trend in the rest of the data, but it should be excluded when drawing the fit line if appropriate.
Remember: A scatter diagram shows how two quantities relate, not that one causes the other; best‑fit lines predict only within the plotted range—extrapolation is uncertain.
Scatter graphs and correlation Stretch
Use and interpret scatter graphs of bivariate data; recognise the direction (positive/negative), strength (weak/strong) and whether a trend exists; draw an estimated line of best fit to make predictions by interpolation or extrapolation, keeping in mind that correlation does not prove causation.
Key words
positive correlation
as one variable increases, so does the other
How to do it
Examine how the points spread on the x‑y plane.
Decide whether they rise (positive) or fall (negative) overall.
Judge tightness of clustering to assess strength: tightly clustered = strong, loosely scattered = weak or none.
Sketch a straight line that most closely follows the pattern – this is your line of best fit.
Use the equation of the line (slope m and intercept c) or read directly from the graph to interpolate within the data range or extrapolate beyond it.
State direction, strength, and note that any inference is only about correlation, not causation.
Watch out for
Mis‑reading axes scales; ensure units match.
Drawing a line through the origin without checking data spread;
Using interpolation outside the plotted range; avoid extrapolation unless clearly justified.
Remember: Correlation tells only that two variables vary together, never why they do so. Always report direction, strength and note any outliers before drawing conclusions.
✏️ Try 3 quick questions
Which of the following points has the largest \(y\)-value? (1,2), (2,5), (3,7). Give that value.
How many different pairs (x,y) can be plotted from the following sets of numbers such that each x is paired with each y exactly once?
X: 3, 7, 11
Y: –2, 5
(Answer with a single integer.)
The temperature (°C) at altitudes of 0 km, 1 km and 2 km are 15, –5 and –25 respectively. Assuming a straight line trend, what is the predicted temperature at an altitude of 3 km?
There are more on this in 4 Triangles, with hints and worked answers, and real past-paper questions. Carry on there.
Practise it with real exam questions
4 Triangles has quick questions on scatter graphs and correlation, real GCSE past-paper questions with their mark schemes, and a revision plan that picks what to work on next.
Questions on this topic for a lesson, homework or revision - a quick "do now", misconception questions, small steps that change one thing at a time, and practice, with answers on the last page.