CHARTS / PRACTICAL GUIDE
Scatter plot: examples and when to use it
A scatter plot places each observation at a pair of numeric coordinates. It shows whether two measures vary together, whether the relationship is curved, and which observations stand apart.
When to use it
Use it for paired measurements such as dose and yield, price and demand, or two performance indicators. A fitted line can summarize an approximately linear relationship.
When to choose another view
Do not use it for a category and a single value: a bar chart is clearer. Do not reduce a curved relationship to one correlation coefficient without inspecting the plot.
How to read it
An upward pattern suggests a positive relationship; a downward pattern suggests a negative one. Look at spread, clusters and outliers. A visible relationship alone does not establish a causal effect.
Data format and example
Two numeric columns, with each row representing the same observation in both columns. Optional labels identify observations; a third measure can size bubbles.
| Dose | Yield |
|---|---|
| 10 | 21 |
| 20 | 29 |
| 30 | 33 |
| 40 | 41 |
| 50 | 46 |
Make it in Datamon
- Keep X and Y paired by row. Check missing values and confirm that both columns are numeric.
- Select Scatter and map X and Y. Include units in the axis titles.
- Inspect outliers before adding a fitted line. Check whether the relationship looks linear.
- Use the correlation and regression tools for a reproducible calculation, then add the result to your explanation.
Common mistakes to check
- Dropping a missing value from one column without its paired value shifts the observations and changes the answer.
- Outliers can dominate a fitted line. Investigate them; do not silently delete them.
- Correlation describes linear association, not causation, and a small value can hide a nonlinear pattern.
Calculate the relationship with the Pearson correlation calculator or fit a line with the regression calculator.