// Traditional Research Lines
Our laboratory explores the fundamental capabilities and limitations of human cognition to inform the design of complex systems. 1 > Our lab improves cyber hygiene compliance through usable designs and deeper psychological understanding. 2 > We strive to predict effective and efficient interfaces by applying our knowledge of visual search information processing. 3 > We are exploring the effectiveness of situated AI notifications on end-user cybersecurity decision-making.
01. Human-Centered Cybersecurity
Our lab explores the psychological drivers behind cybersecurity behaviors (e.g., cyber hygiene; Cain, Edwards, & Still, 2018 Notebook Podcast | Error Classification; Grandhi & Still, 2025b; Notebook Podcast | Personality profiling; Grandhi & Still, 2025a). We investigate graphical authentication schemes that improve usability while maintaining high security. Our efforts have resulted in innovative graphical authentication prototypes (e.g., Cain & Still, 2017; Tiller, Angelini, Leibner, & Still, 2019) and their associated provisional patents. To bridge the gap between research and practice, we have provided a list of authentication design guidelines (Still, Cain, & Schuster, 2017) and consolidated the recognized usability issues into usability strengths/weaknesses (Cain & Still, 2018; Mator, et al., 2020). In addition, we have explored the over-the-shoulder-attack vector from a behavioral perspective (e.g., Cain, Werner, & Still, 2017; Cain, Chiu, Santiago, & Still, 2016), a recognized weakness of graphical authentication. We examined the impact authentication schemes have on our limited working memory resources (Still & Cain, 2019; Mator & Still, 2025;Notepad LM Podcast). Recently, we have been developing a Cognition-Inspired Design framework (e.g., Still & Still, 2025). Recently, we have been examining user susceptibility to SMiShing (SMS Phishing) to develop countermeasures (e.g., Edwards et al., 2026; Edwards & Still, 2026).
02. Predicting Attentional Deployment
We explore the influence of saliency within complex stimuli and reveal an array of unanswered questions about the applied boundary conditions of saliency. There is clear value in being able to predict and direct attention with saliency within interfaces (Still & Masciocchi, 2010; Masciocchi & Still, 2013; Still & Still, 2019), but current research offers limited guidance to interface designers (Still, 2017). We utilize computational models and eye-tracking data to predict varying levels of search efficiency (Still & Still, 2024; Still & Still, 2026). Notably, we have focused on mobile-specific web page convention map development (Still & Hicks, 2020; NotebookLM Plus Podcast) and other specific design patterns (Still, 2025). Our future goal is to help designers structure information hierarchically so that critical elements are noticed immediately.
03. Human-Centered AI
We have been keeping pace with the always-evolving landscape of cybersecurity (e.g., 2FA (Still & Tiller, 2021), AI automated driving (Garcia, et al., 2022), SMiShing (Edwards et al., 2023)). We are attempting to bridge the communication gap in cyber hygiene. This effort was recently funded by the Commonwealth Cyber Initiative (CCI). Users' ability to recognize, comprehend, and anticipate the outcomes of cybersecurity threats is known as Cyber Situation Awareness (CSA) (Mator & Still, 2021). Every day users not only demonstrate poor CSA, but also a poor understanding of how to respond to cyber threats. We are exploring a range of AI design solutions to address this societal need (e.g., Katsarakes & Still, 2026).