Abstract: Emphasis effects – visual changes that make certain elements more
prominent – are commonly used in information visualization to draw
the user’s attention or to indicate importance. Although theoretical
frameworks of emphasis exist (that link visually diverse emphasis
effects through the idea of visual prominence compared to background
elements), most metrics for predicting how emphasis effects
will be perceived by users come from abstract models of human
vision which may not apply to visualization design. In particular,
it is difficult for designers to know, when designing a visualization,
how different emphasis effects will compare and how to ensure that
the user’s experience with one effect will be similar to that with
another. To address this gap, we carried out two studies that provide
empirical evidence about how users perceive different emphasis effects,
using three visual variables (colour, size, and blur/focus) and
eight strength levels. Results from gaze tracking, mouse clicks, and
subjective responses in our first study show that there are significant
differences between different kinds of effects and between levels.
Our second study tested the effects in realistic visualizations taken
from the MASSVIS dataset, and saw similar results. We developed
a simple predictive model from the data in our first study, and used
it to predict the results in the second; the model was accurate, with
high correlations between predictions and real values. Our studies
and empirical models provide new information for designers who
want to understand how emphasis effects will be perceived by users.
TL;DR: Our studies and empirical models provide valuable new information for designers who want to understand and control how emphasis effects will be perceived by users.
Keywords: Human-centered computing, Visualization, Visualization techniques, Perception, Visualization design and evaluation methods
5 Replies
Loading