Lesson 5 of 8 · 13 min
Explanatory variables and scatterplots
Identify the explanatory and response variables for two numerical variables, and construct and read a scatterplot with the explanatory variable on the horizontal axis.
What you will be able to do
- Identify the explanatory and response variables for two numerical variables.
- Construct a scatterplot with correctly placed, scaled and labelled axes.
- Read individual points from a scatterplot and spot points that don't fit the pattern.
Two numerical variables, like the distance of a trip and its fare, don't fall into a handful of categories, so a table of counts won't help. Instead, plot each individual as a point: the explanatory variable along the horizontal axis and the response variable up the vertical axis. The pattern the points make shows the association.
Each point is one trip. Drag a point well away from the others and see how a single wrong value stands out from the pattern.
Definition
Scatterplot
A scatterplot shows bivariate numerical data with one point for each individual. The explanatory variable goes on the horizontal () axis and the response variable on the vertical () axis.
The explanatory variable goes along the bottom because it is the one you know (or set) first, and the one you'll use to predict the other. Reading a scatterplot then works like reading any graph: go across to a value of the explanatory variable, then up to see the response.
A road-safety study measures the speed of a car when the driver brakes and the distance the car needs to stop. Which variable goes on the horizontal axis of the scatterplot?
Constructing a scatterplot by hand
Identify the explanatory and response variables
The explanatory variable goes on the horizontal axis, the response variable on the vertical axis.
Choose a scale for each axis
Cover the smallest to the largest value with equal steps that are easy to read (1, 2, 5, 10 …). An axis may start above zero, but then the scale must still be even.
Label both axes with the variable and its unit
For example, "Distance (km)" and "Fare ($)".
Plot each pair as one point
Go across to the explanatory value, then up to the response value. Mark it with a clear dot or cross.
Check
Count the points against the table, and look again at any point that seems out of place.
Worked example
Construct a scatterplot from a table
A university student recorded the distance and the fare for eight rideshare trips around Brisbane.
| Distance (km) | 2.4 | 3.8 | 5.0 | 6.5 | 8.2 | 9.0 | 11.5 | 14.0 |
|---|---|---|---|---|---|---|---|---|
| Fare ($) | 11.20 | 13.90 | 15.10 | 19.30 | 21.80 | 22.40 | 27.90 | 32.50 |
Construct a scatterplot of the data.
Step 1: Identify the explanatory and response variables
The fare depends on the distance, so distance is the explanatory variable (horizontal axis) and fare is the response variable (vertical axis).
Step 2: Choose the scales
Distances run from 2.4 to 14.0 km, so use 0 to 15 km in steps of 1 km. Fares run from $11.20 to $32.50, so use $0 to $35 in steps of $5.
Even steps that cover all the data, with easy-to-read numbers, make the points quick and accurate to plot.
Step 3: Label the axes
Write "Distance (km)" under the horizontal axis and "Fare ($)" beside the vertical axis.
Step 4: Plot the points
For the first trip, go across to 2.4 and up to 11.20 and mark the point . Repeat for the other seven trips.
Step 5: Check
There should be eight points. Here they rise steadily from left to right, and none looks out of place.
Try it yourself first, then check the working one step at a time.
Answer
A scatterplot with distance (0 to 15 km) on the horizontal axis, fare ($0 to $35) on the vertical axis, both axes labelled with units, and eight points rising from to , as shown below.
Method
Explanatory variable on the horizontal axis, response variable on the vertical axis.
Choose even scales that cover the data.
Label both axes with the variable and its unit.
Plot each pair as one point, then count the points.
Watch out for
Putting fare on the horizontal axis because its numbers are bigger.
Using an uneven scale to squeeze the data in.
The finished scatterplot: distance on the horizontal axis, fare on the vertical axis, every axis labelled with its unit.
Common mistake
Construction errors that cost marks
- The explanatory and response variables on the wrong axes.
- Uneven scales, such as 0, 5, 10, 20, 50 — every step must be the same size.
- Axis labels without units, or no labels at all.
- Points plotted carelessly, such as 22.4 drawn as 24.
Exam tip
How QCAA marks a scatterplot
QCAA marking guides give marks for correctly identifying the explanatory and response variables (and putting them on the right axes), for formatting the scatterplot with appropriate scales and labelled axes, and for accurately plotting the points. If you misplot a point in the exam, cross it out and replot it — you don't need to start the scatterplot again.
A scatterplot also lets you read the data back. Each point is one individual, and points that sit away from the overall pattern deserve a second look: they might be unusual individuals, or mistakes in the data.
The scatterplot shows the hours per week that 12 students spent studying for a subject and their exam marks. How many students studied for more than 8 hours per week?
Summary
A scatterplot shows two numerical variables with one point per individual.
The explanatory variable goes on the horizontal axis and the response variable on the vertical axis.
Use even scales and label both axes with the variable and its unit.
Read a point by going across to the explanatory value and up to the response value; look twice at points that don't fit the pattern.
Finished the lesson?
Try the 2 quick checks above and the lesson completes itself, or mark it complete now.