Research Projects
Overview
We are constantly inundated with information, and the rise of emerging technologies intensifies this visual overload. Complex visualizations, multiple monitor workstations, and emerging technologies employing augmented reality to superimpose information into our real-world views present new challenges in determining what, where, and how to display information. For instance, the next generation of in-vehicle displays aims to project content onto car windshields, providing greater access to that information, and next steps involve placing mixed reality content into the actual visual roadway. However, such approaches pose challenges in managing limited attentional resources. Placing information incorrectly may obscure roadway hazards, potentially overlooked by drivers, while pushing information to the visual periphery increases the likelihood of missing crucial details. Ideal designs should carefully consider optimal information presentation not only in driving contexts but also in other safety-critical environments like physicians monitoring patients during surgery or soldiers using augmented-reality displays to detect hostile targets on the battlefield.
My research explores how to improve visual perception and optimize attention for everyday and safety-critical tasks through display design. To do this, I employ basic and applied research methods, with a core focus on theories of attention and visual perception to address applied questions about optimal display design. Fundamental to my research program is understanding how we can exploit properties of the visual system to improve display design, the trade-off between accessing information and display capacity on attentional resources, and how to effectively design automation aids to guide attention to critical information, all with the consideration that how we see and interact with information will continue to change as future technologies advance.
Ensemble Perception
Imagine a meadow of grass varying in shades of green. As a gust of wind rustles through the field, you can effortlessly process the average green hue of the grass and the general direction the grass is swaying, without having to focus on individual details. In visual perception, this is called ensemble perception, where the visual system automatically and accurately computes summary statistics, like the mean and variability, across a set of similar items. By understanding the mechanisms of this process, we can exploit the properties of the visual system to improve comprehension of basic and complex visualizations. A line of my research examines the application of ensemble perception to data visualizations of temperature trends (Warden, Witt, & Szafir, 2022a), medical images (Witt, Warden, et al., 2022), and hurricane trajectories (Witt, Labe, Warden, & Clegg, 2023). This work suggests that practical applications of ensemble perception to data visualization is a viable design option to improve understanding.
Overlay Clutter
Imagine searching a computer desktop for a specific file. How quickly you can find a file depends on how many items overlay the desktop area. This becomes a greater issue with advanced displays that superimpose information on the real world, like smart glasses or windshields displaying head-up information. As newer technologies embed more information in the real world, it becomes more crucial to quantifying how much information can be displayed. Little work has modeled the influence of clutter for displays overlaying information in the real world. My line of research investigating the impact of display clutter has shown that the ideal display depends on the task (Warden et al., under review; Warden et al., 2023b): when multiple items need to be combined, keeping items close together is useful. But too much information in one space hinders searching for only one item. These findings provide insights for more advanced displays, like smart glasses, that overlay information onto the real world. This line of research continues to model when and how to directly overlay information, with a focus on emerging technologies.
Information Access Effort
Information processing is influenced by its location on a display. This is evident in scenarios like driving, where ease of access influences our information-seeking behavior, such as relying on mirrors rather than turning our heads to check blind spots. This common failure to check the blind spot highlights a tendency to seek less effortful sources of information over those more challenging to access. Such an issue becomes more problematic and pervasive as information displays continue to advance from basic analog dials to multiple monitors or emerging technologies that project virtual content all around the user. These examples illustrate my research examining the impacts of information access effort, or how much mental or physical effort is needed to access information. This research explores the trade-off between positioning information close together or further apart depending on information priority and how information is used (Warden et al., 2023a; Warden et al., 2022b; Poole, Warden, et al., 2023). The goal is to ensure easy access to information for safety-critical tasks, such as driving, in the face of advancing technologies.
Attention Guidance
Optimizing attention allocation in real-world scenarios is crucial, especially for time-sensitive tasks, such as guiding a driver's attention to roadway hazards. My research delves into the design and effectiveness of automated attention cueing aids, exploring factors like cue placement, clutter, and information access. Early work (Warden et al., 2022) revealed that minimal, centrally located cues with high realism were the most effective in real-world visual searches when cues were correct, but this cue suffered the most when the cue was wrong leading to a high degree of automation bias (Warden et al., 2023). These findings prompted my broader exploration into attention guidance to explore design features that best serve static and dynamic search tasks. My ongoing work seeks to understand the nuanced relationship between information access, clutter, cue precision, and dual-cueing, and how these design features impact imperfect automation. This work addresses critical gaps in cue design for automated attention guidance, particularly in dynamic search, contributing valuable insights to this unexplored field.