Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

More Skin, More Likes! Measuring Child Exposure and User Engagement on TikTok

Miriam Schirmer, Angelina Voggenreiter, Jürgen Pfeffer

TL;DR

This paper quantifies child exposure on TikTok by analyzing 5,896 videos and 432,178 English-language comments from 115 accounts featuring under-13 children managed by adults. Using manual video annotation and BERTopic-based topic modeling, it identifies distributions across Family, Fashion, and Sports categories, examines appearance- and safety-related comments, and assesses how exposure influences engagement and content diffusion. Key findings include that 19.57% of videos show exposed clothing and 3.73% involve makeup, with exposure amplifying appearance-focused feedback and concerns while increasing likes but reducing downloads; a subset of content is removed from or copied to other platforms, raising privacy and safety concerns. The results highlight significant risks associated with sharenting, underscore gendered patterns of exposure, and advocate for education and stronger policy/regulatory measures to protect children's privacy and well-being online.

Abstract

Sharenting, the practice of parents sharing content about their children on social media platforms, has become increasingly common, raising concerns about children's privacy and safety online. This study investigates children's exposure on TikTok, offering a detailed examination of the platform's content and associated comments. Analyzing 432,178 comments across 5,896 videos from 115 user accounts featuring children, we categorize content into Family, Fashion, and Sports. Our analysis highlights potential risks, such as inappropriate comments or contact offers, with a focus on appearance-based comments. Notably, 21% of comments relate to visual appearance. Additionally, 19.57% of videos depict children in revealing clothing, such as swimwear or bare midriffs, attracting significantly more appearance-based comments and likes than videos featuring fully clothed children, although this trend does not extend to downloads. These findings underscore the need for heightened awareness and protective measures to safeguard children's privacy and well-being in the digital age.

More Skin, More Likes! Measuring Child Exposure and User Engagement on TikTok

TL;DR

This paper quantifies child exposure on TikTok by analyzing 5,896 videos and 432,178 English-language comments from 115 accounts featuring under-13 children managed by adults. Using manual video annotation and BERTopic-based topic modeling, it identifies distributions across Family, Fashion, and Sports categories, examines appearance- and safety-related comments, and assesses how exposure influences engagement and content diffusion. Key findings include that 19.57% of videos show exposed clothing and 3.73% involve makeup, with exposure amplifying appearance-focused feedback and concerns while increasing likes but reducing downloads; a subset of content is removed from or copied to other platforms, raising privacy and safety concerns. The results highlight significant risks associated with sharenting, underscore gendered patterns of exposure, and advocate for education and stronger policy/regulatory measures to protect children's privacy and well-being online.

Abstract

Sharenting, the practice of parents sharing content about their children on social media platforms, has become increasingly common, raising concerns about children's privacy and safety online. This study investigates children's exposure on TikTok, offering a detailed examination of the platform's content and associated comments. Analyzing 432,178 comments across 5,896 videos from 115 user accounts featuring children, we categorize content into Family, Fashion, and Sports. Our analysis highlights potential risks, such as inappropriate comments or contact offers, with a focus on appearance-based comments. Notably, 21% of comments relate to visual appearance. Additionally, 19.57% of videos depict children in revealing clothing, such as swimwear or bare midriffs, attracting significantly more appearance-based comments and likes than videos featuring fully clothed children, although this trend does not extend to downloads. These findings underscore the need for heightened awareness and protective measures to safeguard children's privacy and well-being in the digital age.
Paper Structure (29 sections, 5 figures, 4 tables)

This paper contains 29 sections, 5 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Children on TikTok
  3. Sharenting
  4. Sexualization of Children on TikTok
  5. Child Protection Mechanisms on Social Media
  6. Scope and Research Questions
  7. Methods
  8. Data
  9. Recurrent Themes
  10. Video analysis
  11. Video Annotation
  12. Comment Classification
  13. Topic Modeling
  14. Results
  15. Comment Analysis
  16. ...and 14 more sections

Figures (5)

  • Figure 1: Example preview of child-related video content from each category (left to right: Fashion, Sports, Family).
  • Figure 2: Overview of most common appearance-related words in our dataset.
  • Figure 3: Most frequent appearance-based words per category.
  • Figure 4: Overview of topics and their most salient words (selected topics based on their coherence).
  • Figure 5: Intertopic Distance Map: Visualization of the relationships between topics, based on their embeddings, displayed in a 2D space.