Charts
Scatter Plot Maker
Plot pairs of values, with or without a line of best fit — the first thing to draw for two numeric variables.
Have a spreadsheet? Drop a CSV here
Or . CSV, TSV or plain text. You pick which column is which after it loads — the names in your file do not have to match ours. The file is read in your browser and never uploaded.
How to read this result
The scatter plot is the check that every correlation and regression needs. Anscombe's quartet is four data sets with identical means, variances, correlations and regression lines — one linear, one curved, one driven by a single point, one a vertical stack. The summary statistics cannot tell them apart and the picture separates them instantly.
Look for four things. Direction: does y rise or fall with x? Form: straight or curved? Strength: tight or diffuse? And outliers: any point far from the pattern, especially one far out along x, which can pivot the whole fitted line.
The line of best fit is least squares. It minimises the sum of squared vertical distances to the points. It is drawn for convenience; for the standard errors, p-values and residual plot that make it interpretable, use linear regression.
Overplotting hides density. Past a few hundred points, dots start to pile up and the visual impression stops matching the data. Smaller markers, transparency or a heatmap of the 2-D density handle that better.
The formula
The least-squares line. r and R² are reported beside it; for inference on the slope, see linear regression.
Worked example
Ten pairs with a clear linear relationship
Ten observations of x against y — the same data as the linear regression example.
x: 1 to 10 · y: 2.1, 4.3, 6.2, 8.9, 10.1, 12.8, 14.2, 16.7, 18.1, 20.4
- Plot each pair as a point.ten points, rising left to right
- The correlation between the two columns.r = 0.9987
- Squared, that is the share of variance the line explains.R² = 0.9975
- The fitted line.y = 0.30 + 2.015x
A correlation of 0.999 with ten points looks conclusive, and the scatter plot is what lets you confirm it is a straight line rather than a curve that happens to fit.
Checked against R's cor and lm.
The calculator above is loaded with these numbers by the Load the worked example button.
Assumptions, and when to use something else
A scatter plot makes no statistical assumptions — it is a drawing. The assumptions arrive with the fitted line, which is least squares and inherits everything linear regression assumes.
The practical limits are perceptual: overplotting at large n, and the difficulty of judging density from dots. Both are handled by transparency, smaller markers or a two-dimensional density plot.
- You want the full statistical outputLinear regressionStandard errors, p-values, prediction intervals and a residual plot.
- You only want the strengthCorrelation coefficientPearson, Spearman and Kendall together.
- The x axis is timeLine graph makerJoining points in order is right for a sequence, wrong for a relationship.
- There are too many points to seeHeatmap makerA grid of counts handles density where dots cannot.
Questions people ask
What does a scatter plot show?
The relationship between two numeric variables: its direction, its form (straight or curved), its strength, and any points that do not fit the pattern. It is the first chart to draw for paired data and the last check before quoting a correlation.
Should I add a line of best fit?
Add it when the relationship is plausibly linear and you intend to talk about the slope. Adding one to a curved relationship is misleading — the line will look reasonable and describe nothing.
What does the correlation on the chart mean?
Pearson's r, between −1 and 1, measuring linear association. See correlation coefficient for how to read it and why Spearman's version is worth checking alongside.
How do I download the chart?
Every chart draws to a canvas in your browser. Use the download button for a PNG, or right-click the image and save it. Nothing is uploaded to produce it.