Task-Driven Evaluation of Aggregation in Time Series Visualization

Abstract

Many visualization tasks require the viewer to make judgments about aggregate properties of data. Recent work has shown that viewers can perform such tasks effectively, for example to efficiently compare the maximums or means over ranges of data. However, this work also shows that such effectiveness depends on the designs of the displays. In this paper, we explore this relationship between aggregation task and visualization design to provide guidance on matching tasks with designs. We combine prior results from perceptual science and graphical perception to suggest a set of design variables that influence performance on various aggregate comparison tasks. We describe how choices in these variables can lead to designs that are matched to particular tasks. We use these variables to assess a set of eight different designs, predicting how they will support a set of six aggregate time series comparison tasks. A crowd-sourced evaluation confirms these predictions. These results not only provide evidence for how the specific visualizations support various tasks, but also suggest using the identified design variables as a tool for designing visualizations well suited for various types of tasks.

Keywords: Information visualization; perceptual study; time series visualization; visualization design.

Figure 1

We can infer how well a particular encoding support a given task by examining the interplay of visual variables (what visual channels are used to encode value), mapping variables (which raw or derived quantities are visualized), and computational variables (how these quantities are computed).

Figure 2

Visual designs explored in this experiment. The first two rows of encodings use position to encode value; the bottom two use color. Conditions 2d, 2b, 2c, 2g, 2f, and 2h calculate and display different statistics at the per-month scale, which requires prior task knowledge ( e.g. that the tasks will be performed at the scale of months).

Figure 3

We consider the design variables of a visualization in order to make predictions about how it supports different aggregate comparison tasks. We analyzed 8 time series visualization techniques using 3 variables, considering how each variable aligns with task requirements to hypothesize about their performance for 6 tasks. Blue squares indicate the variable aligns with the task, red show misalignments, and grey indicate no prediction.

Figure 4

A summary of our experimental results. All measures are in accuracy across all participants. Gray rows indicate position encodings; white indicate color encodings. Gray columns indicate summary comparison tasks; white columns indicate point comparison tasks. An ”X” indicates that the encoding does not afford that task. and so no experiment was conducted for this combination of task and encoding. Since performance is not strictly comparable across tasks, cell color encodes the number and direction of standard deviations from the task mean: −1,(-0.5,-1),[0.5,-0.5],(1,0.5), ≥1.