How to Make a Scatter Plot on Mac & Windows

A scatter plot is how you see the relationship between two numbers — whether they move together, how tightly, and which points break the pattern. Here's how to make and read one on Mac & Windows.

A scatter plot puts one numeric variable on the X axis and another on the Y, with a dot per observation. It's the first thing to reach for when you ask “do these two things move together?”

Make it

Open your dataset and choose the Scatter chart.
Pick the X and Y numeric variables.
Optionally color the points by a category, or size them by a third variable (a bubble chart).

A scatter plot of two numeric variables in Stratum — Two numbers, one dot per row.

Add a line of best fit

Turn on a trend line to summarize the relationship: a straight ordinary-least-squares fit for a linear trend, or a flexible LOESS curve when the pattern bends. The line makes the direction and strength obvious and is the top thing people add to a scatter plot.

A scatter plot with an OLS line of best fit in Stratum — A best-fit line summarizes the relationship.

Describe it: direction, form, strength, outliers

Four things describe any scatter plot:

Direction — do the points rise (positive) or fall (negative) left to right?
Form — is the pattern straight (linear) or curved?
Strength — tight to the line, or a loose cloud?
Outliers — points that break the pattern. Click one and Stratum reveals that row in the grid.

Correlation isn't causation. A tight upward cloud means the two variables move together — not that one causes the other. We unpack that in the next post.

For the full walk-through, see Lesson 11.

Download Stratum

Frequently asked questions

How do I add a line of best fit to a scatter plot?

Switch on the trend line. Choose an ordinary-least-squares line for a straight relationship, or a LOESS curve when the pattern bends. It summarizes the direction and strength at a glance.

What are the four things to describe a scatter plot?

Direction (positive or negative), form (linear or curved), strength (tight or loose), and outliers (points that break the pattern).

How can you tell if a scatter plot shows positive or negative correlation?

If the cloud of points rises from left to right, the correlation is positive; if it falls, it's negative; if there's no slope, there's little or no linear correlation.

Can a scatter plot be used for categorical data?

No — a scatter plot needs two numeric variables. For categories, use a bar chart, mosaic, or box plot by group instead.