Left, Right, and Gender: Exploring Interaction Traces to Mitigate Human Biases

Understanding How In-Visualization Provenance Can Support Trade-Off Analysis

“…Most recently Lumos [19], [20] added to scatterplots (among other views) in-situ visualizations of interaction traces: real-time ones, such as coloring individual data-points or attributes to encode frequency of interaction; and summative ones to show the distribution of data-points that users interacted with compared to a target distribution on a dimension. Their goal was to enhance awareness of biases [19] and mitigate them [20] during exploration and decision making. In a series of laboratory and crowdsource experiments, they found that these traces increased awareness of biases [19], and somewhat mitigated unintentional bias but enabled intentional bias [20].…”

“…Their goal was to enhance awareness of biases [19] and mitigate them [20] during exploration and decision making. In a series of laboratory and crowdsource experiments, they found that these traces increased awareness of biases [19], and somewhat mitigated unintentional bias but enabled intentional bias [20].…”

Undoing Seamlessness: Exploring Seams for Critical Visualization

“…Vancisin et al [70] propose provenance-driven visualization to represent the levels of transformations in historical documents and make visible the labor, interpretation, and curation of historical documents and data. Wall et al [72] visualize interaction traces by highlighting and summarizing the data points that a user has interacted with to raise "awareness of potential unconscious biases. " Feng et al [31] contribute a similar system, offering cognitive support to change exploration behavior and "[encourage] people to visit more data and recall different insights after interaction. "…”

The Influence of Visual Provenance Representations on Strategies in a Collaborative Hand-off Data Analysis Scenario

“…There is a need for accurate techniques that compress temporal events that occur while matching an appropriate level of temporal granularity and summarization to best serve different audiences. Some techniques focus on preserving the timing and order of events to allow for the review of specific analysis turning points (i.e., History representations) [17,42,75] while others provide a high-level summary of topics reviewed and remove elements of timing completely to make it easy to see what has been explored and what needs further analysis (i.e., Coverage representations) [20,57,68]. Provenance helps collaborators to maintain common ground as they work synchronously [14], or asynchronously [72,75].…”

“…While the potential value of provenance information is strong, core challenges remain with how to process provenance data and design effective representations to support easy human understanding. As has been well documented in the visualization community, the representation of information can have dramatic effects on human interpretation of data [16,20,64,68]. Furthermore, there is relatively limited empirical knowledge of how provenance information is used in hand-off scenarios where a second analyst continues an analysis with provenance records from a prior analyst [75].…”