Offsetting Perceptual Bias in Visual Clustering: The Role of Point Size Adjustment in Variable Display Sizes

TL;DR

This paper investigates how scatterplot size across different displays biases visual clustering and tests whether adjusting point size can offset this perceptual bias. It uses a two-stage user study: a preliminary experiment identifies how size alters clustering perception, and a main experiment assesses whether dynamic point-size adjustments can offset the bias across six steps. The findings show that larger scatterplots increase perceived cluster counts and decrease density, while point-size adjustments can substantially mitigate these effects, though residual bias remains, motivating a predictive calibration model. The work has practical implications for cross-device visual analytics and informs the design of responsive visual encodings to maintain consistent cluster interpretations.

Abstract

Scatterplots are frequently shared across different displays in collaborative and communicative visual analytics. However, variations in displays diversify scatterplot sizes. Such variations can influence the perception of clustering patterns, introducing potential biases leading to misinterpretations in cluster analysis. In this research, we explore how scatterplot size affects cluster assignment and investigate how we can offset such bias. We first conduct a controlled study asking participants to perform visual clustering on scatterplots of varying sizes. We found that changes in scatterplot size significantly alter cluster perception in three key features. In our subsequent experiment, we examine how adjusting point sizes can mitigate this bias. As a result, we verify that adjusting point size can effectively counteract the perceptual biases caused by varying scatterplot sizes. We wrap up our research by discussing the necessity and applicability of our findings in realworld applications.

Figures (4)

• Figure 1: Illustration of our research motivation: the perceptual bias in visual clustering due to scatterplot sizes. In communicative or collaborative analytics environments (e.g., classrooms), users often view the same scatterplots on different displays. These displays present scatterplots in various sizes, which can lead to different interpretations in visual clustering analysis.
• Figure 2: Cluster assignment on small (50px) vs large (350px) scatterplots made by the participants of our user study. Smaller scatterplots lead participants to identify less number of clusters with lower density and bigger cluster area.
• Figure 3: The cumulative and individual step success rates in our main experiment. An increase from each step number indicates that the point change successfully offset the perceptual bias resulting from changes in scatterplot size.
• Figure 4: Mean error rate for failed attempts to offset perceptual bias in our main experiment. The decrease in mean error rates indicates that the perceived cluster count progressively converged towards the initial cluster count. The line denotes the 95% confidence interval.