Abstract

Social media platforms increasingly rely on AI-driven recommendation algorithms to curate user feeds, raising questions about transparency and user control. Major platforms now offer “Why am I seeing this?” explanations, yet we know little about whether users find these features genuinely helpful. Through a vignette-based experiment with 120 participants, we compared user perceptions of explanation approaches across Facebook, Instagram, TikTok, and X/Twitter. Participants evaluated explanations on understandability, actionability, clarity, trust, satisfaction, and skepticism.

Users with greater algorithmic literacy rated explanations more favorably than those with less understanding, and this effect outweighed any differences between platforms or explanation styles. Paradoxically, while participants found explanations generally satisfactory, many remained skeptical that platforms fully disclose how their algorithms work. Our findings suggest that better interface design cannot solve transparency problems alone—users need sufficient algorithmic literacy to interpret explanations meaningfully. We discuss implications for platform design, digital literacy education, and user empowerment in algorithmically mediated environments.

Download

Share

COinS