What is it about?
We tend to underestimate the level of correlation (relatedness) present in scatterplots. Here, we change the sizes of some of the points using a function relating to their distance from the regression line. We test the use of these functions in an online experiment, and find that their use makes people significantly more accurate in estimating correlation from scatterplots.
Featured Image
Photo by Markus Winkler on Unsplash
Why is it important?
Scatterplots are often used to inform decision making and communicate science to the public. They are commonly utilised data-driven policy-making, including in healthcare. Nevertheless, people's perceptions of correlation from scatterplots are inaccurate. We thus attempt to correct for the bias in the present work.
Read the Original
This page is a summary of: Adjusting Point Size to Facilitate More Accurate Correlation Perception in Scatterplots, October 2023, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/visxvision60716.2023.00006.
You can read the full text:
Contributors
The following have contributed to this page
