Political Endorsements by TikTok Influencers: Parasocial Interaction, Identification, and Support for Political Candidates
This study examines political persuasion on TikTok by testing how influencers’ direct address and media source labels shape audience responses to a political endorsement video. Using a preregistered 2 × 2 online survey experiment with 486 U.S. adults, we found that participants exposed to direct address, defined as the influencer looking directly at the camera, reported significantly higher identification with the influencer and stronger parasocial interaction than those exposed to non-direct address. Drawing on the Entertainment Overcoming Resistance Model, mediation analyses further showed that direct address indirectly increased intention to support the endorsed political candidate through multiple relational and cognitive pathways, including identification, parasocial interaction, counterarguing, perceived norms, and perceived video credibility. By contrast, media source labels did not significantly affect perceived video credibility, did not indirectly influence candidate support, and did not moderate the effects of direct address. These findings point to the importance of relational cues in short-form political influencer communication. By identifying the mechanisms through which influencer content affects attitudes and behavioral intentions, this study advances theory on entertainment-based persuasion and provides practical insights into political communication and influencer politics in short-form media environments.
Authors
- Zicheng Cheng (ORCID: https://orcid.org/0000-0003-0859-0538)
- Homero Gil de Zúñiga (ORCID: https://orcid.org/0000-0002-4187-3604)
Institutions
- Pennsylvania State University (US)
- University of Arizona (US)
- Universidad de Salamanca (ES)
- Universidad Diego Portales (CL)
Publication Details
- Journal
- Media Psychology
- Published
- 2026-09-24
- DOI
- https://doi.org/10.1080/15213269.2026.2737363
- Primary Topic
- Social Media and Politics
- Type
- article
- Field-Weighted Citation Impact
- 0.00