Start Date

11-8-2016

Description

Security risks often occur because insiders fail to react appropriately to security warnings, due to inattention to the warnings. This study extends security warning design research that has investigated the impact of different designs, including different symbols of threat such as yellow triangles and exclamation marks. This work uses media naturalness theory in an attempt to boost user engagement with security warnings. We integrated validated images of facial expressions depicting fear and disgust, which signaled an environmental threat, into a browser security warning. An fMRI study (N=23) revealed activity located in the right amygdala to be differentially associated among warnings with integrated expressions of fear, disgust, and neutral emotions compared to faceless stimuli. Behavioral measures of response time and self-reported attention were also supportive of the hypotheses. We also propose a follow-up field study using Mechanical Turk to corroborate the fMRI findings. Our work has implications for research and practice.

Share

COinS
 
Aug 11th, 12:00 AM

Integrating Facial Cues of Threat into Security Warnings – An fMRI and Field Study

Security risks often occur because insiders fail to react appropriately to security warnings, due to inattention to the warnings. This study extends security warning design research that has investigated the impact of different designs, including different symbols of threat such as yellow triangles and exclamation marks. This work uses media naturalness theory in an attempt to boost user engagement with security warnings. We integrated validated images of facial expressions depicting fear and disgust, which signaled an environmental threat, into a browser security warning. An fMRI study (N=23) revealed activity located in the right amygdala to be differentially associated among warnings with integrated expressions of fear, disgust, and neutral emotions compared to faceless stimuli. Behavioral measures of response time and self-reported attention were also supportive of the hypotheses. We also propose a follow-up field study using Mechanical Turk to corroborate the fMRI findings. Our work has implications for research and practice.