What Is a Scatter Plot?

Written by Coursera Staff • Updated on

A scatter plot is a type of graph used to show the relationship between two variables. Learn more about how scatter plots work, who uses them, and why.

[Featured Image] A data scientist sits at a desk and uses scatter plots to find relationships between two variables in a data set.

A scatter plot (also called a scatter chart, scatter graph, or scattergram) allows you to visualize relationships between two variables. Its name comes from the graph's design—it looks like a collection of dots scattered across an x- and y-axis. Some scatter plots have a regression line to indicate the statistical strength of the relationship. Others use multiple colors to distinguish the points.

Anyone working with numbers has likely worked with a scatter plot at some point in their career. This includes scientists, economists, and researchers. Environmentalists and meteorologists also may use scatter plots to indicate variables like temperature or precipitation.

Types of scatter plots

The two types of scatter plots are two-dimensional (2D) and three-dimensional (3D). As the names indicate, a 2D scatter plot has coordinates on a two-dimensional graph with an x- and y-axis (like you see on a piece of paper). A 3D scatter plot uses a three-dimensional grid system incorporating a z-axis to show additional features like scale, product size, or price.

What is a scatter plot used for?

You may use a scatter plot to visualize the relationship between variables, especially when you're looking for trends and want to make predictions. For example, you want to show a connection between consuming sugar and weight. You could plot this on a scatter plot, with one axis representing the amount of sugar each person consumed during a specific time period. The other axis may represent the pounds gained or lost during that time.

Representing the data in this format helps you analyze it. You can see how the different data points relate to each other.

Identify correlations.

When viewing data on a scatter plot, you can usually see correlations, a measurement of the relationship between the data points. A positive correlation means both data points increase. In the example of the relationship between sugar consumption and weight, you may notice that people who consumed more sugar also gained more weight. On the scatter plot, the dots would start in the chart's bottom left corner and move toward the chart's upper right corner.

A negative correlation means one data point increases while the other decreases. When this happens, the points on the scatter plot move from the upper left corner toward the lower right corner. If the variables do not correlate, the points on the scatter plot have no definable shape. A curvilinear relationship appears on the scatter plot as a curve, suggesting a variable is not moving at a constant rate. 

Make predictions.

After identifying the correlations between the variables, you can make predictions. If the scatter plot shows a positive correlation between sugar consumption and weight gain, you can predict that you will lose weight if you reduce your sugar intake. Similarly, you may decide to adjust your marketing budget if you notice a strong positive correlation between sales and engagement on a specific channel. 

Examples of when to use a scatter plot

Scatter plots are more useful when you're working with paired numerical data and have no more than two variables to examine. Consider using a scatter plot when you want to do the following:

  • Explore possible causes and effects. Say you want to investigate the relationship between the types of content—videos, blog posts, tutorials, etc.—you're posting on your website and sales. You could collect data on sales revenue and the web pages that referred customers to complete the sale. Organizing the data in a scatter plot can help you identify which content leads to sales, and you can adjust your content strategy accordingly.

  • Test for autocorrelation. If you notice the number of website visitors is highest over the weekend and lowest on Monday, you could use a scatter plot to look for trends in the traffic pattern over time. In this case, you may compare the number of visitors on Sunday and Monday for six weeks. If the pattern continues—website traffic peaks on Sunday and drops on Monday—you may choose to change the type of content you post on Monday.

  • Examining the relationship between two variables. A human resources director may use a scatter plot to explore the connection between employees' salary and their sense of job satisfaction. After administering a job satisfaction survey, they could plot the survey results along with the salary of each participating employee and look for the connection between the two.

Who uses scatter plots?

People who use data to draw conclusions and make predictions are more likely to use scatter plots. This includes those who work with numbers—economists, project managers, and scientists. If you are a data journalist, market analyst, or researcher, you may use scatter plots to identify trends and predict how people will behave.

Pros and cons of using scatter plots

Although scatter plots help show the relationship between variables, they are one of many ways to visualize data. Understanding the pros and cons of this type of graph can help you decide if it's the right tool for your project.

Here are some pros and cons of using scatter plots.

Pros of using scatter plots

  • Scatter plots are easy to read. You typically can identify a correlation with a glance.

  • Scatter plots can show non-linear relationships. Some data may show up along a curved line or an irregular formation.

  • Scatter plots are easy to create. You can draw one by hand or create one in a computer program like Excel.

  • Scatter plots identify correlation. Knowing the relationship between the variables is a starting point for additional analysis.

Cons of using scatter plots

  • Scatter plots allow for limited analysis. You can use a scatter plot to visualize two—sometimes three—variables, so you need another method for additional analysis.

  • Scatter plots do not indicate causation. Correlation is not the same as causation—two variables can be positively or negatively related and caused by additional variables that may not be indicated on the scatter plot.

  • Scatter plots with too many data points may be difficult to read. Overlapping data can dilute the data and slow the analysis process.

How to get started using scatter plots

You can manually draw a scatter plot or create one in a program like Excel, Tableau, Visme, and Canva. If you're learning how to use a scatter plot, drawing one by hand can help you understand how it works. To start, gather your data and record it in a two-column chart. If you want to create the scatter plot by hand, draw a graph and assign a variable to the x-axis and the y-axis. For each data set, place a dot on the spot where the two values intersect on the graph.

Getting started with Coursera

Scatter plots are one way to visualize data. If your career plans include jobs that require data analysis, you may find it beneficial to learn more about the different ways to present data and draw conclusions from it. Consider a course like Introduction to Data Analysis Using Excel from Rice University or Data Visualization with Python from IBM. These courses include sections on creating scatter plots and other tools for data visualization. Both courses are available on Coursera.

Keep reading

Updated on
Written by:

Editorial Team

Coursera’s editorial team is comprised of highly experienced professional editors, writers, and fact...

This content has been made available for informational purposes only. Learners are advised to conduct additional research to ensure that courses and other credentials pursued meet their personal, professional, and financial goals.