Classroom Activities and New Classroom Apps for Enhancing Children's Understanding of Social Media Mechanisms
Henriikka Vartiainen, Nicolas Pope, Juho Kahila, Sonsoles López-Pernas, Matti Tedre
TL;DR
This study investigates how to improve 5th and 8th graders' understanding of social media mechanisms and their data agency through a four-hour, technology-enhanced intervention in 12 Finnish classrooms. It employs two classroom apps (a profiling game and Somekone) across two workshops, using Likert-based data agency measures and open-ended reasoning tasks analyzed with qualitative content analysis and paired t-tests. Results show significant gains in data agency and in students' data-driven explanations of data traces, profiling, and engagement, with notable shifts from naive folk theories to more scientific conceptions (e.g., advanced profiling/recommendation reasoning increased to $d=0.63$-level effects in several measures). The work demonstrates that tightly designed collaborative inquiry with intentionally visualized data processes can foster conceptual change about opaque social media mechanisms, offering practical implications for computing education design and teacher professional development in data literacy and AI-related topics.
Abstract
Young people are increasingly exposed to adverse effects of data-driven profiling, recommending, and manipulation on social media platforms, most of them without adequate understanding of the mechanisms that drive these platforms. In the context of computing education, educating learners about mechanisms and data practices of social media may improve young learners' data agency, digital literacy, and understanding how their digital services work. A four-hour technology -- supported intervention was designed and implemented in 12 schools involving 209 5th and 8th grade learners. Two new classroom apps were developed to support the classroom activities. Using Likert-scale questions borrowed from a data agency questionnaire and open-ended questions that mapped learners' data-driven reasoning on social media phenomena, this article shows significant improvement between pre- and post-tests in learners' data agency and data-driven explanations of social media mechanisms. Results present an example of improving young learners' understanding of social media mechanisms.
