Escaping from the echo chamber: understanding user behavior from the perspective of psychological reactance theory
Purpose The widespread use of recommendation algorithms offers convenience to users but may unintentionally create echo chambers. This study aims to investigate user behavior in response to echo chambers and the psychological mechanisms that drive these reactions. Design/methodology/approach Based on psychological reactance theory, we develop a four-stage model consisting of “external stimulus – threat perception – reactance – behavioral response” to investigate how echo chambers affect users’ psychology and behavior on social media. This study collects 493 valid questionnaires. Findings Findings suggest that echo chamber features – high information density and similarity – significantly motivate users’ psychological resistance, with perceived freedom threat and fatigue experiences mediating this process. In response to this resistance, users interrupt their behavior, limit self-disclosure and increase misrepresentative interactions. Originality/value This study systematically investigates the psychological and behavioral mechanisms behind users’ reactions to the echo chamber effect. It innovatively applies psychological reactance theory to social media echo chambers, enhancing the understanding of user psychology in the digital era. Additionally, the study broadens the previous emphasis on passive avoidance by identifying a range of coping strategies—both off-platform and in-platform—and emphasizes how users utilize algorithmic logic to modify information flows. Highlights
Institutions
- Qinghai University (CN)
- Northwestern Polytechnical University (CN)
- Huazhong Agricultural University (CN)
- Harbin Institute of Technology (CN)
- Northwestern Polytechnic University (US)
Publication Details
- Journal
- Information Technology and People
- Published
- 2025-08-11
- DOI
- https://doi.org/10.1108/itp-08-2024-0984
- Citations
- 5
- Primary Topic
- Geochemistry and Geologic Mapping
- Type
- article
- Field-Weighted Citation Impact
- 12.40