Effectiveness of fake news inoculation workshop on TikTok behaviour

In this project, we investigate to what extent a Fake News workshop – in which students actively learn about Fake News – is effective in changing students’ TikTok behavior. Firstly, the goal is to collect TikTok data from students who have attended the workshop. Secondly, we aim to demonstrate a causal relationship between attending the workshop and behavioral changes on TikTok.

To collect data from students who have participated in the workshop, a class appointment is scheduled with participating teachers and students after the workshop. Subsequently, TikTok data is i) requested; ii) downloaded; and iii) transferred to a secure data environment in the classroom together with the students and teachers. Secondly, the data is analyzed using statistical models that can demonstrate a causal relationship between the workshop and TikTok behavior. For this purpose, control and treatment groups are created, comparing students who attended the workshop (treatment group) with students who have not yet attended the workshop (control group).

We have collected data from 55 students across 6 different classes who have participated in the workshop. This results in a total of 4.6 million data points – information about which videos are watched by students at what time. The general differences in TikTok behavior among students are significant. For example, some students browse or share many more videos than others. The impact evaluation shows that the workshop had no effect on TikTok behavior. Specifically, we find that the workshop did not lead to students i) browsing less; ii) sharing less; iii) liking less; and iv) watching videos for longer or shorter durations. Follow-up research may delve deeper into the content of the videos and analyze whether behaviour has changed concerning videos that may contain Fake News.