'Treat me as your friend, not a number in your database': Co-designing with Children to Cope with Datafication Online

Title of the Paper

"Treat me as your friend, not a number in your database: Co-designing with Children to Cope with Datafication Online"

Paper Information

• Subject Areas: Human-Computer Interaction (HCI), Algorithm Transparency, Data Privacy Protection, Children's Online Datafication
• Keywords: Datafication, Data Inference, Online Platforms, Children, Co-design

Research Background and Issues

• Identified Issues:

  • Children's behavioral data is systematically recorded, tracked, aggregated, and analyzed on online platforms, and used for behavior prediction, content recommendation, and commercial operations (e.g., targeted advertising).
  • Current datafication practices often lack transparency, including how data is collected, used, and inferred, raising concerns about privacy and autonomy for users, including children.
  • Most children have limited understanding of datafication, and adults also lack sufficient knowledge on this subject.
  • Existing research indicates that while children have a basic understanding of datafication practices, their cognitive and social developmental stages limit their ability to fully grasp the implications of data inference.

• Significance:

  • Children are central users in the digital world and are significantly impacted by datafication, yet their ability to address data privacy issues is limited and requires design support.
  • This issue directly relates to children's digital rights, personal autonomy, and the broader ethical environment in a datafied society.

• Research Motivation:

  • To expand the understanding of children's experiences with online datafication practices and explore their needs and platform design solutions.
  • To provide age-appropriate design support to help children better understand datafication processes and enhance their ability to manage data inference.

• Research Questions:

  • RQ1: How do children perceive datafication practices, and how do they wish to be supported?
  • RQ2: What design mechanisms can help children better cope with datafication?

Solutions

• Methods and Solutions:

  • Using co-design, 10 design activities were conducted, including:
    • Pre-design Activities: Exploring individual understanding of video recommendation and personalized advertising mechanisms.
    • Co-design Activity 1: Using fictional scenarios to guide children in discussing the impact of datafication practices on themselves.
    • Co-design Activity 2: Leveraging the CAL (Critical Algorithmic Literacy) framework to help children design and adjust platform transparency and control mechanisms.
  • Platform Selection: YouTube was chosen as a case study to explore its recommendation system and personalized advertising mechanisms.

• Innovations:

  • Age-specific design: Customized design support mechanisms tailored to children's cognitive development stages from ages 7 to 14.
  • Introducing the CAL framework into children's co-design to guide discussions on key mechanisms of datafication.
  • Designing prototypes compatible with transparency and control mechanisms ("EXPLAIN!" and "CONTROL!" cards) to inspire children to propose improvement suggestions.

• Implementation Steps and Key Techniques:

  • Establishing an equal design team of children and researchers, using role-playing, story discussions, and user interface prototype improvements to guide children in expressing their potential needs.
  • Designing datafication mechanisms based on three cognitive frameworks (cognitive, contextual, critical), covering transparency and control aspects.

Research Findings

• Specific Findings:

  • Identified three cognitive stages of children's understanding of datafication practices:
    • Ages 7-9: Basic understanding of data collection concepts but limited comprehension of data sharing or inference.
    • Ages 10-11: Beginning to recognize cross-platform data sharing and personalized models but with limited depth in understanding inference.
    • Ages 12-14: Generally acknowledge the ethical risks of datafication and propose "reform demands" at both technical and societal levels, such as the concept of data decentralization.
  • Children proposed design mechanisms based on respect, transparency, and autonomy, including algorithm transparency, optimized user control, and "humanized" platform interactions.

• Advantages and Contributions:

  • Provided a systematic understanding of datafication design from children's perspectives.
  • Emphasized respect for humanity and the ethics of addressing children's digital rights, contributing to the creation of child-friendly datafication mechanisms.
  • Proposed specific design directions supporting "data autonomy" and "digital data decentralization."

• Experimental Results:

  • Children of different age groups demonstrated varying needs for algorithm transparency and control, highlighting the importance of non-uniform support methods.
  • Children desired transparent information expression but felt overwhelmed by complex technical terminology, preferring simple and direct user interfaces with friendly interactions.

• Limitations and Future Directions:

  • Limitations:
    • The use of fictional stories may conceptually influence the research results.
    • The study primarily explored children's needs through co-design activities; future work should validate design solutions in real platform scenarios.
  • Future Directions:
    • Develop platform prototypes to validate age-specific design mechanisms.
    • Support children's online data autonomy and the development of related skills.
    • Combine ethics and design to guide platforms toward a "more humanized" and "autonomy-supportive" direction.