Track 3 · Visualising and Exploring Data · Lesson 6 of 11
Scatter plots
Read a scatter plot for direction, form, strength and unusual points, use a fitted line to make a prediction, and fix overplotting.
A property analyst wants to know how much an extra square metre adds to a home's price; a marketing manager wants to know whether TV advertising sells juice. Both start the same way: plot one quantity against the other and look before computing anything.
A scatter plot shows two numeric variables measured on the same cases: one point per case, the explanatory variable on the horizontal axis () and the response on the vertical axis (). Describe what you see in four words:
- Direction: do values tend to rise (positive) or fall (negative) as increases?
- Form: does the cloud follow a straight line, a curve, or separate clusters?
- Strength: how tightly do the points hug that form?
- Unusual points: individual cases far from the pattern.
Included with a Formula Quest membership
It has 8 graded exercises with feedback, hints and worked solutions. After it you will be able to:
- Describe a scatter plot by direction, form, strength and unusual points
- Use a fitted line to predict and compute residuals, staying within the range of the data
- Recognise curved patterns that Pearson's r understates, and confounding that a scatter cannot resolve
- Fix overplotting with jitter, transparency, sizing by count or binning