Spring 2025 Student Research
Research Team 1
Miranda Duffy | Farheen Saiyed | Haley Salas | Angelika Tokarczyk
Misinformation on social media shapes beliefs, behaviors, and policies. Platforms use various warning messages to curb “fake news,” but their effects on trust and engagement are unclear. Our study examines how different moderation approaches—platform-driven, community-driven, and none—affect users’ unconscious and conscious responses. Through eye-tracking, GSR, facial expression analysis, and self-report surveys, we assess how these responses influence trust in moderation and willingness to engage with flagged content. Based on existing literature, we developed a series of hypotheses centered on expectations of a more positive impact of community-driven (compared to platform-based) misinformation moderation on trust, visual attention, emotional engagement, valence, and arousal. The study was conducted with 39 participants using a within-subject experimental design. All participants were exposed to three misinformation moderation conditions (no alert, platform-driven message, community-driven message) using two content topics (healthcare and politics) in a counterbalanced order. Our findings reveal that misinformation moderation intervention types do impact users' trust, visual attention, and physiological responses, while underscoring a critical need to be skeptical of moderation's unintended impacts, such as censorship and the erosion of trust in platforms. Our research contributes to the conversation on how misinformation moderation should focus on equipping users to be skeptical and autonomously discern content for truth or falsehoods. We find support that transparency and crowd-sourced warning messages (community-driven moderation) encourage more autonomous decision-making and resilient user behavior against misinformation.
Research Team 2
Ryan Hensley | Mesones Mesones | Patryk Suszko | Jake Wilcox
Notification Dynamics: An Investigation into Eye Tracking and Galvanic Skin Response to Examine Contextual Influences on User Engagement and Memory Recall
This study investigated the impact of digital notification interruptions on user focus, stress levels, and task performance in a simulated office environment. With the rise of workplace communication tools such as Microsoft Teams and Slack, employees are increasingly subject to interruptions that may influence productivity and cognitive load. To explore these effects, participants completed an urgent email response task that involved synthesizing information on stock levels, shipping times, and pricing. The task was performed under three distinct conditions: no notifications, auditory notifications, and combined auditory-visual notifications. Throughout the experiment, eye-tracking data and galvanic skin response (GSR) were collected to assess changes in visual attention, physiological stress, and memory recall. Results showed that notifications presented in unexpected locations—specifically the top-left corner of the screen—received significantly higher fixation counts, indicating increased visual attention.
However, participants also spent less time reading these notifications, suggesting they were more likely to dismiss them quickly. In contrast, notifications appearing in the bottom-right corner, the standard position for most office platforms, resulted in more consistent engagement and less disruption. These findings suggest that both the modality and placement of digital notifications
can influence user interaction and cognitive efficiency. For organizations aiming to balance productivity with effective communication, maintaining standard notification placements may reduce disruption while preserving message clarity. Future research should replicate this study in real-world office environments and investigate additional variables, such as notification urgency, alternative communication platforms, and varying user roles within workplace settings.
.jpg)
Research Team 3
Stella Burns | Grace Jakobs | Lauren Devine | Caroline Makara
Sh*t You Won’t Forget: The Surprising Link Between Memory, Subconscious Responses and Word Choices in the Context of Video Lectures
This study investigates the controversial and understudied impact of profanity in academic video lectures on student engagement, physiological arousal, and information recall. While profanity is traditionally considered inappropriate in educational contexts (Jay & Janschewitz, 2008; Mohamed Saat et al., 2004), emerging literature suggests that taboo language may enhance memory encoding due to its emotional salience and expectancy violation (Burgoon, 1993; Vingerhoets et al., 2013). Despite these claims, existing studies often rely on self-reports and do not adequately examine subconscious and physiological responses. This research fills that gap by leveraging biometric data to measure the effects of profanity on (1) emotional engagement through facial expressions, (2) emotional arousal via galvanic skin response (GSR), and (3) information recall through quiz performance. Thirty-one undergraduate students (aged 20–26) from a private Midwestern university participated in a within-subject experimental study. Each participant viewed three academic video lectures—each identical in content but varying in profanity level: none, low (e.g., “damn”), and high (e.g., “fuck”). Eye tracking (SmartEye AI-X), facial expression recognition (AFFDEX via iMotions), and GSR (Shimmer3) were used to capture biometric responses. After each video, participants completed a 5-question recall quiz containing 3 "affected" questions (tied to profanity sentences) and 2 "unaffected" questions. Emotional engagement was measured via facial muscle activation metrics, including basic emotions and valence. GSR peaks per minute indexed emotional arousal. The experiment included demographic surveys, calibration protocols, and a post-test debrief, including manipulation checks for perceived profanity intensity.
Research Team 4
Ashton Handorf | Miguel Picazo-Marin | Clay Strader
Visual and Physiological Insights into Human Deepfake Detection: an Eye-Tracking and GSR Study
This paper evaluates the ability of someone to detect deepfake media, and biometric responses that occur when a subject is actively trying to discern between deepfake and genuine media. This paper evaluates the ability and speed one is able to detect a deepfake, whether or not humans have a biometric response upon deepfake detection, and their overall concern of this technology's potential impact on society . In this within-subject experimental study, data from 32 Midwestern university students were collected using eye-tracking, galvanic skin response, and surveys. Our study revealed that dwell time was the most statistically significant eye-tracking metric, with participants spending longer fixating on medium- and high-quality deepfakes, suggesting heightened scrutiny due to subtle imperfections. Low-quality deepfakes elicited faster initial fixations but lower dwell counts, implying obvious tells accelerated detection. GSR data showed no significant physiological response to deepfake quality, indicating detection may rely more on cognitive than autonomic processes. Facial deepfakes triggered significantly more fixations than audio manipulations, highlighting visual features as primary cues for inauthenticity
Research Team 5
Shaquan Coombs | Kathryn Glasgow | Nabeha Sajid
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.
​
Research Team 6
Damian Acosta |S hiv Patel | Anel Zukanovic
Marketing in the Macabre: Does a Horror Movie Trailer’s Content Determine Interest Levels
The purpose of the study is to determine how the amount of plot information shown in a horror movie trailer (High vs. Low Spoilers) and the trailers format (Teaser vs. Full-Length) influence viewers’ attention, physiological arousal, emotional reactions, memory for content within trailer, and willingness to watch the movie shown. By combining the various biometric measurements along with self-reported data through surveys, the research aims to generate actionable insights that can help filmmakers and marketers in the film industry create trailers that maximize engagement while minimizing audience fatigue and movie spoilers. In this within-subject experimental study, data from 30 university students were collected utilizing eye-tracking, galvanic skin response, and survey questionnaires.