Scatter plots use an array of dots or points to show an association or correlation between 2 measures (e.g. children’s age and height).
Using scatter plots
Each point in a scatter plot aligns a single data measurement on the x axis with its corresponding measure on the y axis. If there is an association between these 2 measures, the ‘scatter’ of points will form a telling shape. For example, the shape of the scattered points will approximate a diagonal line from the bottom left to the top right of the graph if values along the y axis tend to increase along with values on the x axis. This indicates a positive, linear correlation (see graph below). Stronger correlations are evidenced by a tight concentration of points around this line; a greater spread of points indicates a weak correlation. Scatter plots can also be useful for showing nonlinear (e.g. curvilinear) associations.
Trend lines
A trend line (or ‘line of best fit’) is often added to scatter plots to summarise and highlight the strength of the association between the 2 plotted measures, as shown in the graph above. This strength is clear by both the angle and shape of the line, and the spread of data points around the line. For example, a 45° line indicates a strong, linear correlation; a flat line indicates no association; and a U-shaped line indicates a curvilinear association.
Alternatives to scatter plots
Consider whether your readers will know how to interpret a scatter plot before using one to display correlated data. This type of graph is relatively common in many scientific disciplines, but may be confused with dot plots or line graphs by other audiences.