What Is a Scatter Plot? Read X and Y
Learn how a scatter plot shows correlation, clusters, and outliers, with a worked x-y example you can copy.

A scatter plot places one dot for each observation on a pair of numeric axes. The horizontal position is one measure and the vertical position is another. Nothing connects the dots, because the chart is not a sequence. It is a picture of how two numbers move together, where they clump, and which points sit alone.
In BeCharts you plot those pairs in the scatter plot generator. Paste x and y columns, keep the markers the same size, and export PNG, SVG, or PDF. If a third number should change the circle area, that is a different chart: the bubble chart maker. This article stays with equal dots.
What a scatter plot shows
Each row in the data becomes one point. If you have 40 products, you get 40 dots. Readers then look for three things:
- Direction. Do higher x values tend to sit higher on y, lower on y, or with no visible tilt?
- Strength. Is the cloud tight around a line, or is it a loose spray?
- Exceptions. Which points sit far from the rest?
A scatter plot does not prove that one variable causes the other. Price and units sold can move together because of a third factor you did not plot. The chart only shows the association in the columns you chose.
Parts of the chart
- X axis. The measure you treat as the horizontal variable. In a price-versus-units chart, price often goes here. See x axis and y axis if you are unsure which measure belongs where.
- Y axis. The second measure, drawn vertically.
- Point. One observation. Its coordinates are the two numbers on that row.
- Marker. The shape of the point. Equal size keeps attention on position.
- Series color. Optional. Use a second series when you need to compare two groups on the same axes, such as two regions.
- Axis titles and units. Required if the numbers are not obvious. "Price (USD)" is clearer than "X".
There is no category axis. If your labels are product names with a single value each, use a bar graph maker instead.
A worked example
Suppose a small shop tracks seven products:
| Product | Price | Units sold |
|---|---|---|
| A | 12 | 80 |
| B | 18 | 64 |
| C | 24 | 51 |
| D | 31 | 42 |
| E | 38 | 36 |
| F | 46 | 22 |
| G | 55 | 15 |
Plot price on x and units sold on y. The dots fall from the upper left toward the lower right. Readers can see that higher prices in this table go with fewer units. They cannot see profit, because profit is not on either axis. If margin should be visible as circle size, move the same rows to a bubble chart and add that third column.
The pattern is only as good as the sample. Seven points can suggest a relationship that more products would weaken. Say how many observations the chart includes when you present it.
When to use a scatter plot
Use one when both fields are numbers and the question is about their relationship:
- Price against quantity
- Study hours against test score
- Response time against request size
- Advertising spend against signups, when you are exploring rather than claiming cause
- Two lab measurements on the same samples
Skip it when:
- One field is a category name. Bars compare those.
- The horizontal field is time and you want a trend line through ordered periods. A line graph maker connects those points on purpose.
- You need the share of a total. A pie chart maker answers that, and only for a few slices.
- You need the shape of a single distribution. A histogram maker or a box plot fits better.
How to read the cloud
Positive association. The cloud rises from left to right. Larger x values tend to have larger y values.
Negative association. The cloud falls from left to right, as in the price example.
No clear association. The points form a round cloud or a flat band. Do not force a story onto noise.
Clusters. Two or more tight groups can mean the data mixes populations. Color the groups as separate series before you describe a single relationship.
Outliers. A point far from the cloud can be a data error or the most important observation. Check the source row before you delete it. Zooming the axes or shrinking markers makes distant points easier to see when the middle is crowded.
Overplotting. Hundreds of identical coordinates stack into one blob. Jitter, smaller markers, or opacity helps. The editor includes a jittered scatter template for category-like x values that would otherwise sit on the same vertical line.
Scatter plot versus nearby charts
| Question | Chart |
|---|---|
| How do two numbers move together? | Scatter plot |
| How does one number change over time? | Line chart |
| Which category is largest? | Bar chart |
| How large is a third number at each x-y position? | Bubble chart |
| How are values spread inside one group? | Box plot or histogram |
A line through a scatter plot is a separate analysis step. BeCharts draws the points. It does not compute a regression line or a correlation coefficient. If you need that statistic, calculate it in a spreadsheet and keep the plot as the picture of the raw pairs.
Common mistakes
- Connecting the dots. A line implies an order. Scatter points usually have no order along the page.
- Truncating an axis so a weak tilt looks steep. Start from a scale that includes the real range, and say so if you zoom.
- Encoding too much in color and size at once. Position is the accurate channel. Leave size for a bubble chart, and use color only for a group label.
- Mixing units on one axis. Do not plot dollars and percentages as if they shared a scale.
- Claiming cause. "Higher price reduces units" may be true in an experiment. A shop export only shows that the two columns co-occur.
Build one in BeCharts
- Open the scatter plot generator or a layout in the scatter templates.
- Replace the sample coordinates with your x and y columns. One row is one point.
- Title the axes with units. Add a second series only when the groups should be compared on the same grid.
- Export PNG, SVG, or PDF. No account is required. Charts you save stay in the browser unless you export the file.
If you are still choosing among chart families, start from types of charts and come back here when both fields are numeric.
Frequently asked questions
What is a scatter plot used for?
A scatter plot shows the relationship between two numeric measures. Each dot is one observation, with one measure on the x axis and the other on the y axis.
Does a scatter plot show cause and effect?
No. It shows how the two columns move together in the data you plotted. A cause claim needs a design beyond the chart.
How do I make a scatter plot in BeCharts?
Open the scatter plot generator, replace the x and y values, and export PNG, SVG, or PDF. https://www.becharts.com/scatter-chart-maker
Try in BeCharts
Open ready-made scatter plots templates or jump straight into the editor with sample data loaded.