Research

Technologies
Should Benefit
Everybody

Technology should serve everyone, and that begins with understanding it. I study how humans communicate with artificial intelligence and how to harness it for positive behavior change, without losing sight of who it empowers and who it leaves behind.

Scroll
Jump to

Research Areas

Focus Area

Health & Strategic Communication

Examining message acceptance, behavioral intentions, and the effectiveness of AI-delivered health promotion and advertising messages. How conversational AI agents can promote health behaviors.

Focus Area

AI Trust & Persuasion

How explainability and emotional support shape trust in AI health advisors and persuasive technologies. My dissertation investigates building trust in AI health advisors through explainability and emotional support under Dr. Linwan Wu.

Focus Area

Algorithmic Bias & Digital Inequalities

Investigating who gets left behind: disparities in AI adoption, algorithmic skills, and outcomes across gender, age, and demographic groups.

Methods

Methodological Toolkit

Experimental & Survey

Controlled experiments and surveys to test causal mechanisms in AI perception, trust, and persuasion.

Psychophysiological

Eye tracking, GSR, EEG, and EMG via iMotions to capture real-time cognitive and emotional responses.

Computational & Qualitative

Critical discourse analysis, in-depth interviews, Python-based analysis, and LLM fine-tuning.

Professional Experience

Where I Work

Lab Coordinator · 2025 – 2026

USC Biometrics & User Experience Lab

Coordinating lab scheduling and equipment operations. Training students and researchers in biometric methods including eye tracking, GSR, EEG, and EMG. Supporting campaign effectiveness testing for strategic communication courses and USC's NSAC team.

Lab Analyst · Summers 2023 – 2026

USC Social Media Insights Lab

Producing social media analytics reports for clients including Nephron Pharmaceuticals and Chernoff Newman. Contributing to the methodology and software behind College Sports Social Pulse, featured in Sports Business Journal.

collegesportssocialpulse.com →
Publications

Journal Articles

2026
Perceptions of AI-generated profile pictures: Effects on quality, attractiveness, and trustworthiness Long, J., Xiao, J., McCray, S., Ağaoğlu, E., Al-Ajmi, A., Akalonu, C., & Xu, Y. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 20(4), Article 3

Book Chapters

2025
Artificial intelligence in advertising: Unveiling challenges and opportunities via the consumer's lens Wu, L., Ağaoğlu, E., & Sun, Y. In S. A. Bowen & E. Erzikova (Eds.), Handbook of Innovations in Strategic Communication
2022
Intersection of digital and educational inequalities during the COVID-19 pandemic Ağaoğlu, E. In Condition of Media During Pandemic. UM:AG, Ankara
2021
Elderly's relationship with digital technology education: Micro case of Ankara Municipality Technological Education Center Ağaoğlu, E., Onen, N., & Tokatli, P. In M. Figan & Y. D. Ozdemir (Eds.), Digital Culture, Digital Inequalities and Aging

Conference Presentations

2026
The AI divide: How gender shapes attitudes and skills toward generative AI Ağaoğlu, E., & Al-Ajmi, A. 109th Annual AEJMC Conference, New Orleans, LA
2025
Exploring generational digital inequalities in the age of algorithms Ağaoğlu, E., & Al-Ajmi, A. 108th Annual AEJMC Conference, San Francisco, CA
2025
Exploring the news judgment of large language models Long, J., McCray, S., Ağaoğlu, E., Akalonu, C., & Xiao, J. 108th Annual AEJMC Conference, San Francisco, CA
2025
Perceptions of AI-generated news: An experiment on disclosure type and timing Long, J., Akalonu, C., Xiao, J., Ağaoğlu, E., & McCray, S. 108th Annual AEJMC Conference, San Francisco, CA
2024
AI chatbots for emotional support: Chatbot agency locus and user prior experience Wu, L., & Ağaoğlu, E. 74th Annual ICA Conference, Gold Coast, Australia
2024
Multidimensional investigation of Asian activism: The role of racial esteem, ethnic identity, and bicultural efficacy Jun, J., Kim, J., Woo, B., Zain, A., Xu, Y., Ağaoğlu, E., & Xiao, Y. 74th Annual ICA Conference, Gold Coast, Australia
2024
“Hey AI, write a story for me!”: Investigating gender discourse in AI-powered chatbots through AI-generated stories Ağaoğlu, E. 107th Annual AEJMC Conference, Washington, D.C.
2024
Artificial impressions: Trust and credibility in AI-enhanced profile pictures Long, J., Xiao, Y., McCray, S., Akalonu, C. A., Ağaoğlu, E., Xu, Y., & Alajmi, A. 107th Annual AEJMC Conference, Washington, D.C.
2023
Digital inequalities among the elderly from a cumulative (dis)advantage perspective: A case study from the Anatolian capital Ağaoğlu, E. 106th Annual AEJMC Conference, Washington, D.C.
Work in Progress

Ongoing Research

Pilot
Investigating the impact of language formality on adherence to sexual health promotion messages delivered by conversational AI agents Ağaoğlu, E. Stigma can prevent individuals from seeking sexual-health support. I examine whether an AI chatbot's use of informal language reduces this barrier among college students, a group particularly vulnerable to sexually transmitted infections.
Writing
Digital inequalities in the age of artificial intelligence: Impact of gender Ağaoğlu, E., & Al-Ajmi, A. Through a cross-sectional survey, we examined differences in AI adoption, skills and benefits.
Writing
Algorithmic gender performance: A critical discourse analysis of AI chatbot-generated narratives Ağaoğlu, E., & Cole, A. AI companies pledge to reduce bias in their platforms, and these safeguards work when users explicitly request biased content, but not for implicit bias. Using critical discourse analysis, we examined AI-generated stories produced from neutral prompts, finding that large language models still reflect gendered discourse.
Analysis
Exploring the role of agency locus in the effectiveness of anti-smoking messages delivered by conversational AI agents Ağaoğlu, E., & Mukhtarnejad, M. AI's growing agency creates new tensions between humans and technology. We examine whether framing a chatbot's agency locus as human- or machine-driven affects users' quitting efforts by shaping their sense of relatedness.