Clickbait Defense Pipeline

This work addresses a key gap in clickbait mitigation: platforms deploy warnings, but we lack an evidence-based understanding of how users conceptualize clickbait and why warnings do (or don’t) influence behavior. We conducted a large-scale online study (MTurk, 770 participants) to elicit user mental models of clickbait and evaluate clickbait warnings that convey harm. Results show many users hold simplified mental models that underestimate clickbait’s risks, which helps explain continued susceptibility. We found that harm-focused warnings can support understanding and safer decisions, highlighting the need for interventions aligned with users’ misconceptions. Overall, the paper provides a mental-model lens to design more effective, trust-centered clickbait defenses.

Security education for clickbait is underdeveloped—especially approaches that tailor learning to how users actually think. To bridge this gap, we combined mental models with learning science principles to design education content that corrects misconceptions while keeping users engaged. Across two MTurk studies (834 participants total), we first derived six clickbait mental models, then translated them into a treatment design and compared it to a learning-science-only baseline. Using knowledge measures, UX evaluation, and thematic analysis, we found the mental-model-informed content improved clarity and relevance while supporting stronger learning outcomes. This work offers a principled blueprint for scalable clickbait education that is both usable and psychologically grounded.

Even when users recognize clickbait, curiosity and persuasive framing can still drive risky clicks. This paper tackles that gap by testing persuasion-theory-based interventions (e.g., social consequence, personal consequence, and badge-based designs) that nudge safer behavior. We iteratively refined designs through a preliminary study, then a lab study (20 participants), and finally evaluated them in a large MTurk experiment (773 participants). Findings show persuasion techniques can reduce susceptibility while improving users’ understanding of clickbait. We also derive practical design recommendations—such as varying techniques to avoid habituation and leveraging search/context cues to help resolve curiosity without clicking.