Algorithm-driven Personalization and Digital Addiction--A Field Experiment on Tiktok (Douyin)

Abstract:

This paper investigates how the short-video platforms affect users through a field experiment on Douyin (the Chinese Tiktok). We incentivized the participants to deactivate the personalization system or stop using Douyin for four weeks. Relative to the control, the no personalization group significantly reduces Douyin usage after the intervention and reports higher work/study efficiency as well as more negative attitude toward Douyin. The intervention causes no significant change in usage behavior among the no usage group afterwards, but surprisingly leads to lower mental health status. Cognitive ability and opinion polarization are not changed in both groups. These results enhance our understanding of the effect of short-video platforms and suggest that personalization deactivation can be an effective new method to reduce digital addiction.

Previous post
Trust and Trustworthiness--Experiments with Artificial Intelligence (AI) Agents Next post