Identifying Symbolic Modeling of Grooming Precursors in Social Media

Short-form video platforms such as TikTok and Instagram Reels increasingly circulate dating and lifestyle advice that may normalize coercive relational scripts, framing jealousy as devotion or secrecy as intimacy. A key challenge for preventing technology-facilitated child sexual exploitation is that early grooming precursors can be communicated through audiovisual choices, such as tone, pacing, setting, and editing, that make manipulative narratives appear benign. This qualitative multimedia content analysis identified recurring observable patterns of symbolic modeling through which relational and boundary-shaping cues associated with early-stage sexualized grooming are framed as ordinary dating or lifestyle advice in public TikTok and Instagram posts. Guided by Bandura’s social cognitive theory of mass communication, a purposive sample of 100 archival posts (50 per platform) was coded in MAXQDA using a structured multimodal observation protocol. Beyond its findings, the study contributes a transferable protocol for coding relational risk across visual, audio, and on-screen text channels simultaneously, extending grooming-detection methods beyond their text-only focus. Analysis produced a taxonomy of 10 recurring masking categories, derived from 289 operationally defined codes and interpreted through four themes. Masking operated primarily through presentation-layer conventions: simulated closeness, ordinary platform packaging, staged intimacy, and instructional framing dominated the corpus. Explicit grooming-literature cues were rare, appearing in 19 of 100 posts, and two cues, secrecy normalization and isolation, never appeared. Findings may inform prevention education and support future multimodal detection research.