A scatter graph plots pairs of data values as individual points on a grid. Each point represents one item (e.g. one person) with two measured variables (e.g. their height and shoe size). By looking at the overall pattern of points, we can describe the correlation — whether the two variables tend to increase or decrease together.
Each plotted point = one data item. Two axes = two variables. Pattern of points = correlation.
Example 1 — Identifying correlation
Step 1
Look at the overall pattern of points — ignore individual outliers.
Step 2
If points trend upward from left to right → positive correlation. If they trend downward → negative correlation. If no clear pattern → no correlation.
Step 3
You can also describe strength: points close to a straight line = strong correlation. Points spread widely = weak correlation.
Example 2 — Describing correlation in context
Step 1
A scatter graph plots hours of sunshine (x) against number of ice creams sold (y).
Step 2
Points trend upward — more sunshine, more ice creams sold.
Step 3
Write: "There is a positive correlation between hours of sunshine and ice cream sales — as sunshine increases, ice cream sales tend to increase."
Exam language: always say "positive/negative/no correlation" and give a brief reason using the context of the graph. Do not just say "the points go up".
✓ Quick Check 1 — Scatter graphs and types of correlation
Question 1 of 10
Drawing Scatter Graphs and the Line of Best Fit
To draw a scatter graph, plot each pair of values as a point — one value on each axis. To draw the line of best fit, draw a single straight line through the middle of the data so that roughly equal numbers of points lie on each side of the line. The line does not need to pass through any specific point.
Line of best fit: straight line through the middle of the data — equal points either side
Example 3 — Drawing a scatter graph and line of best fit
Draw axes: x-axis "Hours revised" (0–10), y-axis "Score (%)" (0–100). Label with units and add a title.
Step 2
Plot each pair: (1,30), (2,40), etc. Use a cross (×) or small dot for each point.
Step 3
Draw a straight line of best fit with a ruler — position it so roughly half the points are above and half below. It does not have to pass through any particular point.
Step 4
The pattern shows positive correlation — more hours revised tends to give a higher score.
Do not join the points like a line graph. The line of best fit is a single straight line through the middle of the data — not a connection of each point in order.
✓ Quick Check 2 — Drawing scatter graphs and lines of best fit
Question 1 of 10
Reading and Interpreting Scatter Graphs
Once you can create a scatter graph, the next skill is to read and interpret one. This means identifying patterns, spotting outliers, recognizing different types of correlation, and making predictions from the data.
Reading a scatter graph: Look for patterns (correlation), identify unusual points (outliers), and use the line of best fit to estimate unknown values.
Example: Reading the Height vs Weight graph
Observe
The points trend upward from left to right — positive correlation.
Describe
Points lie close to the line of best fit — strong positive correlation. As height increases, weight tends to increase.
Spot outliers
One red point is far from the line — a person who is tall (≈200 cm) but lighter than expected (≈50 kg). This is an outlier.
Estimate
For someone 188 cm tall, trace to the line of best fit: approximately 75 kg.
Key skills when reading a scatter graph:
1. Describe the correlation (positive, negative, or none)
2. Comment on strength (strong if points are close to the line, weak if spread out)
3. Identify any outliers (points that don't fit the pattern)
4. Use the line to estimate or predict values
✓ Quick Check 2B — Reading and Interpreting Scatter Graphs
Question 1 of 10
Interactive — Correlation Explorer
Click on the canvas to add points. Drag existing points to move them. Click a point to select it (turns gold), then press Delete or the Remove button to delete it. Watch the line of best fit and Pearson's r update live.
0 points
Pearson's r: —
Add at least 3 points
Practice: Lines of Best Fit
Work out each answer. Type your answer, then click Check.
#
Question
Answer
Result
1
A line of best fit passes through (0,10) and (10,30). What is the gradient?
2
Using y = 2x + 10, find y when x = 5
3
A line of best fit passes through (0,5) and (20,25). What is the gradient?
4
Using y = x + 5, find y when x = 15
5
A line of best fit passes through (2,8) and (6,16). What is the gradient?
6
Using y = 3x, find y when x = 7
7
A scatter graph has 15 points; 3 are anomalies. How many points fit the trend?
8
A line of best fit passes through (0,0) and (5,20). What is the gradient?
9
Using y = 4x, find x when y = 32
10
A line of best fit passes through (0,100) and (10,50). What is the gradient?
⚠ Things to Watch Out For
Confusing correlation types: Positive correlation slopes upward (both increase together), negative correlation slopes downward (one increases as the other decreases), and points scattered randomly show no correlation.
Drawing a bad line of best fit: The line should follow the overall trend of ALL the points, roughly equal numbers above and below it — not simply join the first and last points.
Extrapolating too far: Using the line of best fit to predict values far outside the plotted data range is unreliable — the trend may not continue.
Misreading the axis scale: Check what each gridline represents before reading off values — scales are not always in steps of 1.
Correlation is not causation: Two variables being correlated does not prove one causes the other — there could be another explanation.