Adjusting Point Size to Facilitate More Accurate Correlation Perception in Scatterplots

Gabriel Strain; Andrew J. Stewart; Paul Warren; Caroline Jay

doi:10.6084/m9.figshare.24117987.v2

Adjusting Point Size to Facilitate More Accurate Correlation Perception in Scatterplots

Gabriel Strain, Andrew J. Stewart, Paul Warren, Caroline Jay

Research output: Chapter in Book/Conference proceeding › Conference contribution › peer-review

Abstract

Viewers consistently underestimate correlation in positively correlated scatterplots. We use a novel data point size manipulation to correct for this bias. In a high-powered and fully reproducible study, we demonstrate that decreasing the size of a point on a scatterplot as a function of its distance from the regression line is able to correct for a systematic perceptual bias long present in the literature. We recommend the implementation of our technique when designing scatterplots that aim to communicate positive correlations.

Original language	English
Title of host publication	2023 IEEE Vis X Vision
DOIs	https://doi.org/10.6084/m9.figshare.24117987.v2 https://doi.org/10.1109/visxvision60716.2023.00006
Publication status	Published - 12 Dec 2023

Access to Document

Cite this

@inproceedings{1c1763424bb94882919b3f9a5a7a7e99,

title = "Adjusting Point Size to Facilitate More Accurate Correlation Perception in Scatterplots",

abstract = "Viewers consistently underestimate correlation in positively correlated scatterplots. We use a novel data point size manipulation to correct for this bias. In a high-powered and fully reproducible study, we demonstrate that decreasing the size of a point on a scatterplot as a function of its distance from the regression line is able to correct for a systematic perceptual bias long present in the literature. We recommend the implementation of our technique when designing scatterplots that aim to communicate positive correlations.",

author = "Gabriel Strain and Stewart, {Andrew J.} and Paul Warren and Caroline Jay",

year = "2023",

month = dec,

day = "12",

doi = "10.6084/m9.figshare.24117987.v2",

language = "English",

booktitle = "2023 IEEE Vis X Vision",

}

TY - GEN

T1 - Adjusting Point Size to Facilitate More Accurate Correlation Perception in Scatterplots

AU - Strain, Gabriel

AU - Stewart, Andrew J.

AU - Warren, Paul

AU - Jay, Caroline

PY - 2023/12/12

Y1 - 2023/12/12

N2 - Viewers consistently underestimate correlation in positively correlated scatterplots. We use a novel data point size manipulation to correct for this bias. In a high-powered and fully reproducible study, we demonstrate that decreasing the size of a point on a scatterplot as a function of its distance from the regression line is able to correct for a systematic perceptual bias long present in the literature. We recommend the implementation of our technique when designing scatterplots that aim to communicate positive correlations.

AB - Viewers consistently underestimate correlation in positively correlated scatterplots. We use a novel data point size manipulation to correct for this bias. In a high-powered and fully reproducible study, we demonstrate that decreasing the size of a point on a scatterplot as a function of its distance from the regression line is able to correct for a systematic perceptual bias long present in the literature. We recommend the implementation of our technique when designing scatterplots that aim to communicate positive correlations.

U2 - 10.6084/m9.figshare.24117987.v2

DO - 10.6084/m9.figshare.24117987.v2

M3 - Conference contribution

BT - 2023 IEEE Vis X Vision

ER -

Adjusting Point Size to Facilitate More Accurate Correlation Perception in Scatterplots

Abstract

Access to Document

Fingerprint

Cite this