Eye-Tracking the Storm: The Effect of Variation in Presentation of Visual Risk Information

Forecast and warning information is frequently shared with the public via channels that support both textual message and visual content. The presentation of visual risk messages, such as those that feature geographic uncertainty and increasing levels of threat, vary between offices with scant research to support their chosen presentation style or styles to represent levels of risk or threat impact. A quick search of NWS communication on Twitter will demonstrate the various approaches utilized to communicate risk information to the public, suggesting that inconsistency may differ due to office preference and skill of personnel as well as a lack of best practices built from the research record on human risk information processing.

The risk information seeking and processing (RISP) model allows scholars to investigate how individuals process information about a perceived hazard. Most studies using the RISP model have focused on how individuals process information for slow moving environmental or health hazards. However, less is known about how individuals process information about fast moving meteorological threats that require more immediate decision making. Even less is known about how individuals process visual risk messages. Furthermore, visual communication scholars have found a connection between attention, measured by time, or allocation, dedicated to viewing visual information, leading to information processing. For example, prior studies have connected prior experience with an issue or product to more systematic processing of information; however, no research that we are aware of has investigated the dynamic nature of information processing of visual risk communication in relation to prior experience with a hazard, such as a tornado.

In this paper, we add to the scholarship on risk communication by examining individual processing of visual risk images (maps) during a fictitious tornado threat. Using eye-tracking methods, we assess the relationship between participant’s prior tornado experience, image type, and tornado threat on information processing outcomes and the reduction of information insufficiency.

Drawing upon the RISP model, we identify how individuals process risk information; systematically (i.e., attempts to thoroughly understand or evaluate through careful thinking and intensive reasoning) or heuristically (i.e., the activation of well-learned judgments through cues such as source credibility and visual salience) by examining the allocation of visual attention to visual imagery including maps, symbols, legends, and other visual variables such as message placement and content.

Theoretically, attention allocation plays a role in information processing. Individuals only have a limited amount of mental resources that may be allocated to a visual or message. If a person processes information quickly or superficially, the understanding of the contents and the information is more likely to be trivial in comparison to those who systematically process or critically assess the information (i.e., Kahlor, Dunwoody, Griffin, Neuwirth, & Giese, 2003). Specific features of the visual’s structure may impact attention allocation; however, an individual’s prior experience has also been found to impact cognitive processing (i.e., Griffin, Dunwoody, & Neuwirth, 1999). To evaluate prior hazard experience, this study applies Demuth’s (2018) tornado experience scale to assess the effect of hazard experience on visual risk information processing. Hazard experience with a weather event has been theoretically connected to how a person perceives, assesses, and responds to a risk. More specifically, the person’s prior experience relates to the judgment of the event and the probability of processing the information about the risk and disaster more critically. The connection of prior experience to visual attention allocation will allow us to understand how past hazard experience influences heuristic and systematic processing in the amount of time a participant will fixate on areas within a visual hazard message.

By examining variation in visual risk imagery, over time, with differing audience types, this project will provide empirical evidence to develop a preliminary model of Visual-Risk Information Seeking and Processing and will also support the goals of the weather enterprise to more consistently communicate visual information to individuals at risk. Furthermore, by assessing the risk information processing activities relative to varied formats for visualized forecast and warning information, we will begin to develop a baseline of empirical findings for the effective communication of visual risk information.

References

Demuth, J. L. (2018). Explicating experience: Development of a valid scale of past hazard experience for tornadoes. Risk Analysis. doi: 10.1111/risa.12983

Griffin, R. J., Dunwoody, S., & Neuwirth, K. (1999). Proposed model of the relationship of risk information seeking and processing to the development of preventive behaviors. Environmental research, 80(2), S230-S245. doi:10.1006/enrs.1998.3940.

Kahlor, L., Dunwoody, S., Griffin, R. J., Neuwirth, K., & Giese, J. (2003). Studying heuristic-systematic processing of risk information. Risk Analysis, 23(2), 355-368. doi: 10.1111/1539-6924.00314