What Is a Scatter Diagram? Types, Correlation, and How to Read One

A scatter diagram (also called a scatter plot, scatter chart, or scatter graph) is a chart that places data points on two axes — X and Y — to show whether the variables are related. If the dots form an upward pattern, the correlation is positive. Downward, negative. Random, none. The R-squared value quantifies how strong the pattern is.

Scatter diagrams are one of the most fundamental tools in statistics, quality control, and data analysis. This guide covers what they are, the types of correlation they reveal, and how to make one in seconds with the free scatter diagram maker.

Scatter Diagram: Definition and Purpose

A scatter diagram displays the relationship between two numerical variables by plotting data points on a Cartesian coordinate system. The X axis represents one variable (typically the independent variable or the one you suspect causes changes), and the Y axis represents the other (the dependent variable or the one you suspect is affected).

Each point on the chart represents one observation with paired X and Y values. For example, if you measured the height and weight of 50 people, each person would be one dot — their height on the X axis and their weight on the Y axis.

The purpose is simple: does changing X seem to change Y? If the dots form a visible pattern, the answer might be yes. If they are scattered randomly, the answer is probably no. A trend line and R-squared value formalize what your eye already sees.

Other names for the same chart: scatter plot, scatter chart, scatter graph, XY chart, correlation diagram. Different fields and textbooks use different terms, but they all describe the same visualization.

The Three Types of Scatter Diagram Correlation

1. Positive Correlation

Dots trend upward from left to right. As X increases, Y increases. The trend line has a positive slope. Examples: study hours vs. exam score, temperature vs. electricity usage for air conditioning, experience vs. salary.

2. Negative Correlation

Dots trend downward from left to right. As X increases, Y decreases. The trend line has a negative slope. Examples: car age vs. resale price, altitude vs. air pressure, absences vs. course grade.

3. No Correlation (Zero)

Dots are scattered randomly with no visible pattern. The trend line is nearly flat and R-squared is close to zero. Examples: shoe size vs. IQ, hair color vs. typing speed.

Within positive and negative correlations, there are also degrees of strength. A strong correlation shows dots tightly clustered around the trend line (R-squared above 0.70). A weak correlation shows dots loosely scattered around it (R-squared below 0.40). The visual tightness of the dot cluster matches the R-squared number.

Sell Custom Apparel — We Handle Printing & Free Shipping

How to Read a Scatter Diagram in 4 Steps

1. Look at the overall pattern. Do the dots trend upward, downward, or nowhere? This tells you the direction of correlation.
2. Check the tightness of the cluster. Dots hugging a clear line = strong correlation. Dots loosely scattered in a general direction = weak correlation. Cloud with no direction = no correlation.
3. Read the trend line equation. The slope (m in y = mx + b) quantifies the rate of change. "For every 1 unit increase in X, Y changes by m units."
4. Check R-squared. This single number confirms what your eye sees. Above 0.70 = strong. 0.40–0.70 = moderate. Below 0.40 = weak. Near 0 = none.

Also watch for outliers — dots that sit far from the main cluster. A single outlier can dramatically shift the trend line and inflate or deflate R-squared. If you spot one, investigate whether it is a data entry error, a special case, or a genuine anomaly.

Real-World Scatter Diagram Examples

Education: A teacher plots assignment submission time (days before deadline) against grade. The scatter diagram shows a moderate positive correlation — students who submit earlier tend to score higher. R-squared = 0.45, suggesting other factors (effort, prior knowledge) also matter.

Manufacturing: A quality engineer plots machine RPM against product defect rate. Strong positive correlation (R-squared = 0.82) reveals that higher speeds cause more defects. The factory reduces speed and defect rate drops by 30%.

Healthcare: A researcher plots daily step count against resting blood pressure in 200 patients. Moderate negative correlation (R-squared = 0.38) — more steps, slightly lower blood pressure. The relationship exists but is not the whole picture.

Marketing: A marketing manager plots email open rate against click-through rate across 100 campaigns. Weak positive correlation (R-squared = 0.18). Opens and clicks are somewhat related, but a high open rate does not guarantee clicks — the email content matters more.

In each case, the scatter diagram gives a starting point: "Is there a relationship worth investigating further?" The answer shapes the next step — deeper analysis, a controlled experiment, or moving on to other variables.

Create a Scatter Diagram in Under a Minute

Open the free scatter diagram maker and paste your data. The tool renders the chart with a trend line and R-squared instantly. No account, no software, no data uploaded.

Format your data as X,Y pairs, one per line:

15, 72
22, 85
8, 61
30, 90
12, 68

Or upload a CSV file and pick your columns from the dropdown menus. Both methods produce the same output: an interactive scatter diagram you can customize and download as a PNG image.

For more on what the trend line and R-squared mean, read our line of best fit guide. And if you want to compare free scatter plot tools, see the tools comparison post.

Frequently Asked Questions

What is the difference between a scatter diagram and a scatter plot?

There is no difference. Scatter diagram and scatter plot refer to the same type of chart. The term "scatter diagram" is more common in quality control and project management literature (PMP, Six Sigma). "Scatter plot" is more common in statistics and data science.

Who invented the scatter diagram?

The scatter diagram is generally attributed to John Frederick W. Herschel, who used a version of it in 1833 to study stellar observations. Francis Galton later popularized scatter diagrams in the 1880s for studying human traits and regression toward the mean.

Can a scatter diagram show a curved relationship?

The dots may form a curve, indicating a non-linear relationship. However, a standard scatter diagram with a linear trend line will not capture the curve accurately. The R-squared for the straight line will be low even though the variables are related. Non-linear regression requires specialized tools.

How is a scatter diagram different from a histogram?

A histogram shows the frequency distribution of a single variable (how often each value occurs). A scatter diagram shows the relationship between two variables (how X and Y move together). They answer completely different questions.