How to Make a Scatter Plot for Scientific Papers
A complete guide to creating publication-quality scatter plots. From correlation analysis to journal-ready export.
What is a scientific scatter plot?
A scatter plot is a data visualization that uses dots to represent the values of two continuous variables. Each dot’s position is determined by its x and y values. Scatter plots are used to reveal relationships, correlations, trends, and outliers in experimental data.
Key requirements:
- • Axis labels with units for both x and y
- • Regression line with R² and p-value (if applicable)
- • Distinct markers for different groups
- • Colorblind-friendly palette
- • Journal width (single: 84–90 mm, double: 170–183 mm)
- • 300 DPI minimum for raster export
Step-by-Step Guide
- Prepare your data
Collect paired measurements (x, y) for each data point. Ensure both variables are continuous and measured on the same scale or appropriately normalized.
- Choose axes
X-axis: independent variable. Y-axis: dependent variable. Both axes must have labels with units.
- Plot data points
Place each data point at its (x, y) coordinate. Use consistent marker size. For multiple groups, use distinct markers and colors.
- Add regression line
If showing correlation, fit a linear regression line. Display R² and p-value in the legend. Do not force a line if no correlation exists.
- Add error bars
If applicable, add horizontal and vertical error bars to show measurement uncertainty. State error type in the legend.
- Export for publication
Set width to journal column width. Export at 300 DPI. Use a colorblind-friendly palette. Ensure markers are distinguishable at journal size.
Common Mistakes to Avoid
- ✗Forced regression lines — Only add a regression line if there is a clear correlation. Forcing a line on uncorrelated data is misleading.
- ✗Overlapping markers — Use semi-transparent markers or jitter if points overlap. Overlapping makes the plot unreadable.
- ✗Missing units — Both axes must have labels with units. "Time" is not enough; "Time (days)" is correct.
- ✗Too many groups — Limit to 3–5 groups per plot. More groups become confusing and markers hard to distinguish.
Frequently Asked Questions
How do you make a scatter plot for a scientific paper?
To make a scatter plot: (1) collect paired data (x and y values), (2) plot each data point as a dot, (3) add axis labels with units, (4) include a regression line if showing correlation, (5) add error bars if applicable, (6) use distinct markers for different groups, (7) export at 300 DPI at journal width.
What is a scatter plot used for?
A scatter plot is used to show the relationship between two continuous variables. It reveals correlation, trends, outliers, and clusters. Scatter plots are essential for showing experimental data, clinical correlations, and dose-response relationships.
Should scatter plots include regression lines?
Yes, if you are showing a correlation. Include the regression line with the equation, R² value, and p-value in the figure legend. Do not force a regression line if there is no clear correlation — this can be misleading.
What is the best tool for making scatter plots?
FigureGuild is ideal for publication-ready scatter plots. It auto-fits regression lines, calculates R², and applies journal formatting. R (ggplot2) and Python (matplotlib/seaborn) are also used but require coding. Excel is not recommended for publication due to poor typography and limited formatting.
How do you show multiple groups in a scatter plot?
Use distinct markers (circles, squares, triangles) and colors for different groups. Add a legend clearly identifying each group. Ensure colors are colorblind-friendly. FigureGuild automatically applies distinct marker styles and accessible palettes.
Should scatter plots show error bars?
Error bars on scatter plots are optional. Add them if both x and y variables have uncertainty. Horizontal error bars show x uncertainty; vertical error bars show y uncertainty. If uncertainty is small, error bars may be omitted for clarity.
Related Pages
Create Scatter Plots Automatically
FigureGuild auto-fits regression lines, calculates R², and applies journal formatting — all from your raw data.
Try FigureGuild Free