top of page

Current Projects

2026

STUDENTS, GENERATIVE AI, AND COGNITIVE OUTSOURCING

  • Felipe Bravo, Noah Lepore, Adelisa Pelinkovic, Jillian Rossman; Dinko Bačić; Conference Manuscript Under Review

This ongoing project examines how access to generative AI may be shaping students’ engagement with learning tasks and academic performance. As tools such as ChatGPT become more embedded in educational settings, the study explores whether AI serves primarily as a support for learning or whether it may also reduce the degree of effort students invest in academic work. Using a multimethod approach that combines biometric, behavioral, and self-report data, the project investigates how AI availability may influence attention, cognitive engagement, stress-related responses, and performance across different types of learning activities. The goal is to contribute objective evidence to current conversations about the role of generative AI in higher education and its implications for student learning.

2026

AI PREDICTIONS, TRUST, DECISION OVERRIDE

  • Finn Norman, Favour Mamudu,  Mariana Vargas Diaz, Dinko Bačić; Conference Manuscript Under Review

This ongoing project examines how people respond to AI-generated recommendations in decision-making contexts, with a particular focus on situations where AI outputs conflict with their own expectations. The study explores trust not simply as an attitude, but as something reflected in whether individuals rely on or override AI guidance. More specifically, it investigates how the degree of mismatch between human expectations and AI predictions, along with the quality of accompanying explanations, may shape reliance on AI-supported recommendations.

2026

AI-GENERATED FASHION MEDIA

  • Alaina McClanen-Clemons, Sawyer Smith, Ruby Voke;  Conference Manuscript Under Review

The growing use of generative AI in digital fashion media raises important questions about how consumers perceive and interact with AI-generated imagery compared to human-created content. This study examines differences in visual attention, behavioral engagement, and perceptual responses when individuals view AI-generated versus human-generated fashion images in a social media environment. Using a controlled laboratory experiment, participants will view curated fashion boards resembling Pinterest-style interfaces while their visual attention is recorded using eye-tracking technology. Behavioral engagement will be captured through image “save” (pinning) behavior, complemented by post-exposure perception measures. Drawing on theories of processing fluency, visual attention, and authenticity perceptions, the study proposes that AI-generated imagery may attract greater visual scrutiny due to perceptual irregularities, yet generate lower engagement because users perceive human-created content as more authentic. The research further examines whether the presence of human faces moderates these effects. By integrating biometric attention data with behavioral engagement outcomes, this study contributes to emerging research on AI-generated media, consumer authenticity judgments, and human–AI interaction in visually driven digital platforms.

2026

AI LANGUAGE CONFIDENCE AND ELABORATION

  • Karoline Shipton, Margot Hartley, Aniko Kittridge, Dinko Bačić;  Conference Manuscript Under Review

As artificial intelligence increasingly functions as a cognitive partner in decision-making, understanding how users evaluate and trust AI recommendations becomes critical. While prior research has emphasized algorithmic accuracy, considerably less attention has been given to how the presentation of AI output, particularly linguistic confidence and elaboration, shapes trust independent of correctness. This study proposes a research agenda examining how these communication cues influence user trust, attention, and emotional engagement during AI-assisted decision-making. Drawing on persuasion theory, trust in automation, and the Elaboration Likelihood Model, the research investigates how confidence framing, explanatory detail, and information accuracy interact to shape user responses

2026

AI DISCLOSURE IN DIGITAL ADVERTISING

  • Meghan Economos, Ediz Kerim, Benjamin Liddell, Gavin Murphy;  Conference Manuscript Under Review

As generative AI becomes increasingly embedded in digital advertising production, firms face growing pressure to disclose AI involvement. While prior research shows that AI-origin disclosures often reduce credibility and consumer evaluations, little is known about how disclosure affects consumers during advertisement exposure. This study examines how AI disclosure design (presence, wording, and timing) shapes visual attention, physiological arousal, perceived authenticity, and advertising effectiveness. Across three within-subject laboratory experiments, participants view digital image and video advertisements while eye-tracking, facial expression analysis, and galvanic skin response data are collected alongside survey measures. We test whether disclosure imposes an attentional and authenticity cost that alters ad and brand attitudes. By integrating biometric process measures with evaluative outcomes, this research shifts focus from post-exposure judgments to real-time cognitive and emotional processing, offering theoretical and practical insight into responsible AI transparency in digital media.

2024-2026

COGNITIVE MISFIT & INSIGHT CREATIVITY

  • Tarek Soukieh, Ray Henry, Dinko Bačić, Charles McElroy;  Manuscript Under Review (A* Journal)

In this research project, we explore how cognitive misfit—a mismatch between task and data visualization—can enhance creativity by disrupting habitual thinking and fostering deeper engagement. Unlike cognitive fit, which optimizes for efficiency by reducing cognitive effort, misfit forces users to actively interpret unfamiliar representations, potentially leading to novel insights. This perspective challenges the traditional assumption that alignment between task and visualization always improves decision-making performance. A key theoretical mechanism linking cognitive misfit to creativity is mindfulness, defined as a state of active, present-focused cognitive engagement. While cognitive fit may encourage automatic reliance on established mental models—misfit disrupts these patterns, prompting users to pay closer attention, process information in new ways, and generate more creative solutions. By integrating cognitive misfit and mindfulness into visualization design, this research stream has the potential to advance both IS theory and AI-driven decision-support systems.

2024-2025

INFORMATION OVERLOAD AND DECISION MAKING

PROJECT COMPLETED

In this research stream, we explore how bounded rationality, information overload, and brand familiarity shape consumer decision-making, particularly in high-information environments. We investigate how consumers, constrained by cognitive limitations, process product information and allocate attention when faced with overwhelming choices. Specifically, we examine information overload as a factor that increases cognitive strain, leading individuals to rely on familiar brands to simplify their decisions. We also analyze brand familiarity, focusing on how prior knowledge and brand loyalty impact attention and product preferences. To capture both conscious and subconscious decision-making processes, we utilize biometric measures, including eye-tracking to assess visual attention strategies and Galvanic Skin Response (GSR) to measure physiological arousal. By integrating these methods, this research stream provides a deeper understanding of how consumers navigate complex product choices, offering insights into decision-making behavior, user experience, and the role of cognitive effort in shaping preferences in real-world purchasing environments.

2024-2026

GENDERED MARKETING CUES & VISUAL ATTENTION

  • Anna Skinner, Evey Kallmeyer, Paige Gutierrez, Dinko Bačić; Conference Manuscript Published - Conference on HCI - MIPRO 2025

  • Anna Skinner, Evey Kallmeyer, Paige Gutierrez, Dinko Bačić; Manuscript Under Review 

In this research stream, we examine how gendered colors and endorsers influence subconscious visual attention using eye-tracking technology. We investigate whether traditional gender-based color associations and endorser gender affect where and how consumers allocate their visual attention. Grounded in the Eye-Mind Hypothesis and the SOR (Stimulus-Organism-Response) model, we explore whether gendered marketing cues impact consumer perception in ways that align with societal norms or if subconscious behaviors reveal a more nuanced reality. We analyze fixation duration, dwell time, saccades, and time to first fixation to assess engagement and preference patterns. This research stream contributes to understanding gendered marketing effectiveness, providing insights for product design and advertising strategies. By integrating biometric data, we uncover implicit consumer biases and decision-making processes that may not align with self-reported preferences, offering a more accurate view of gendered consumer behavior.

2024-2026

SUBTITLES - VISUAL ATTENTION  - RECALL

  • Kaitlin Walker, Ayla Lezic, Dinko Bačić; Stage: Conference Manuscript Published - Conference on HCI - MIPRO 2025

  • Kaitlin Walker, Ayla Lezic, Eoin McDonagh, Bilal Khurshid, Dinko Bačić; Manuscript Draft

In this research stream, we explore the impact of subtitles on recall, visual engagement, and arousal in streaming content, guided by dual coding theory. We investigate how subtitles influence comprehension-based recall (understanding of spoken content), perception-based recall (memory of visual scene details), and subconscious visual attention using eye-tracking technology. Additionally, we evaluate how subtitles affect both native and foreign language content, assessing whether their benefits extend beyond language accessibility. To further understand the cognitive and emotional impact, we analyze arousal levels using Galvanic Skin Response (GSR) data, measuring how subtitles influence viewers' physiological engagement. By examining fixation duration, dwell time, gaze allocation, and GSR responses, we assess whether subtitles enhance memory retention, shift attention, or increase cognitive effort. Our findings contribute to the growing body of knowledge on multimodal learning, providing insights for streaming platforms, content creators, and educators to optimize subtitle integration without disrupting engagement.

2023-2026

COGNITIVE FIT THEORY - LITERATURE REVIEW

  • Dinko Bačić, Nenad Jukić; Manuscript Draft

​

Cognitive Fit Theory (CFT) has been a foundational framework for understanding how the alignment between problem-solving tasks and information representations influences performance. Over the past three decades, a substantial body of research has applied and extended CFT across various domains, yet a comprehensive synthesis of this work has been lacking. This structured literature review examines CFT research from 1991 to 2023, tracing its development, identifying key research communities, and analyzing study methodologies. We explore the diverse contexts in which CFT has been applied, including data visualization, decision-making, information systems, and emerging technologies such as artificial intelligence and augmented reality. This review synthesizes existing research and highlights unresolved questions, providing a foundation for advancing CFT scholarship and extending its applicability to contemporary and future technological landscapes.

2025 - 2026

PROFANITY IN VIDEO LECTURES

  • Lauren Devine, Stella Burns, Grace Jakobs, Caroline Makara, Dinko Bačić; Conference Manuscript Published - AMCIS 2025 (Montreal, Canada)

  • Lauren Devine, Stella Burns, Grace Jakobs, Caroline Makara, Dinko Bačić; Manuscript Draft

​

​

In this research stream, we examine the impact of profanity in educational video lectures on recall, visual attention, and emotional engagement. We explore whether the strategic use of swear words enhances memory retention, captures visual focus, and triggers emotional arousal. Profanity is known to activate heightened cognitive processing, potentially improving recall by increasing salience and engagement. Using eye-tracking technology, we assess fixation duration, dwell time, and blink rates to measure shifts in attention. Additionally, we employ Facial Action Coding System (FACS) and Galvanic Skin Response (GSR) to analyze subconscious emotional reactions and physiological arousal. By varying profanity intensity across different conditions, we evaluate how different levels of taboo language influence learning effectiveness and engagement. Our findings contribute to pedagogical strategies and instructional design, challenging traditional views on language appropriateness in education while uncovering new pathways to enhance student engagement.

2025 - 2026

DEMENTIA BEHAVIORAL INTERVENTION

  • Katie Ehlman Nicolas, Dinko Bačić, Nicolas Jankuhn; Poster Presentation Accepted

  • Katie Ehlman, Dinko Bačić, Nicolas Jankuhn, Swateja Nimkar; Manuscript Under Review (Q1)

​

In this research stream, we explore the Positive Physical Approach (PPA) as a behavioral intervention designed to enhance caregiver interactions with people living with dementia (PLWD). Developed by Teepa Snow in 2005, PPA is a 12-step method that focuses on retained cognitive and motor skills rather than deficits, using familiar social cueing to facilitate engagement. To measure intervention fidelity, we employ facial action coding (FACS) and eye-tracking technology, providing objective insights into attention and affective responses. By improving caregiver-PLWD interactions, PPA aims to reduce resistance to care and enhance daily living assistance. Our findings contribute to advancing neurologically-informed caregiving techniques, refining behavioral interventions that prioritize connection and preserved abilities in dementia care.

2025 - 2026

MISINFORMATION MODERATION

​​

In this research project, we examine how misinformation moderation strategies impact user trust, engagement, and cognitive responses on social media. We compare platform-driven moderation (top-down enforcement by social media companies) and community-driven moderation (crowdsourced interventions) to assess their effectiveness in shaping perceptions of content accuracy and trust in moderation itself. Using biometric measures, including eye-tracking, Galvanic Skin Response (GSR), and facial expression analysis, we explore how misinformation warnings affect visual attention, emotional arousal, and willingness to engage with flagged content. By integrating self-reported and physiological data, our study provides deeper insights into how users process and react to misinformation warnings, offering guidance for social media platforms, policymakers, and misinformation researchers in designing effective moderation systems.

2025 - 2026

NOTIFICATION DYNAMICS

​

​

In this project, we investigate the cognitive and emotional impacts of digital notifications on task performance, visual attention, and physiological arousal in high-pressure work environments. Grounded in Multiple Resource Theory and interruption research, we examine how different notification modalities (visual, auditory, combined) and contextual factors such as urgency and placement affect task efficiency, memory recall, and cognitive load. Using eye-tracking technology, we assess fixation duration, gaze patterns, and attention shifts, while Galvanic Skin Response (GSR) measures physiological arousal in response to distractions. By integrating self-reported and biometric data, this research offers insights into designing less disruptive notification systems, balancing workplace efficiency and well-being, and optimizing digital work environments for enhanced focus and productivity.

2025 - 2026

VISUAL DESIGN CHOICES AND CLIMATE URGENCY PERCEPTIONS

This study examines how design elements of data visualizations affect attention and urgency toward climate change. Specifically, we focused on climate crisis data visualizations, adjusting the position of key takeaways and manipulating images to amplify the message. Our experimental conditions enabled us to show how viewers interpreted the climate data visualizations and how viewers engaged with the visualizations. In our study, we utilized eye-tracking technology to gain insights into how participants process information while observing the impact on viewer attitudes via surveys. The outcomes help us recommend the best ways to present climate change visualizations to capture viewers' attention.

2025

DEEPFAKE DETECTION

PROJECT COMPLETED

​

Deepfakes have been prevalent in digital media for at least over 5 years, but their effectiveness at deceiving the audience has not been fully researched. Researchers are well aware that deepfakes mislead a percentage of viewers, but it is still unclear what makes some deepfakes more realistic and other comically obvious. Both the quality and manipulated features of deepfakes have been described as key indicators to expose deepfake media’s authenticity. Regarding manipulated features, extensive studies have documented detection rates of specific features, but have led to conflicting results. Furthermore, detection with differentiating deepfake quality has yet to be tested for significance. This study analyzes which variables lead deepfakes to mislead the audiences’ to believe them as real. Before developing our experiment, we crafted 4 questions that aimed to guide our research: RQ1: How well are people able to detect a deepfake? RQ2: How fast are people able to detect a deepfake? RQ3: What are the features of deepfake media that expose its inauthenticity? RQ4: Is there a biometric response when a subject has detected deepfake content?

bottom of page