The modality-congruent carryover effect: How difficulty in identifying (deep)fake news impacts self-confidence in truth discernment and susceptibility to subsequent disinformation

Summary

Choi reports a two-wave online experiment investigating how the experience of trying to authenticate one piece of (deep)fake news shapes responses to subsequent fake content. Participants encountered celebrity (deep)fake news in either text or video form within a social media context, and the study tracked downstream effects on self-confidence in truth discernment and susceptibility to further disinformation. The central argument is that there is a modality-congruent carryover effect: difficulty authenticating fake news in one format propagates to later fake content, but only when the subsequent material appears in the same modality (text-to-text or video-to-video).

Key Contributions

• Introduces the concept of a modality-congruent carryover effect for sequential (deep)fake news exposure.
• Provides experimental evidence that authentication experiences with one fake item shape vulnerability to later fake items.
• Shifts disinformation research from single-exposure paradigms toward sequential, cumulative exposure dynamics.
• Identifies modality (text vs. video) as a moderator of cumulative disinformation susceptibility.

Methods

• Two-wave online experiment with temporal separation between initial and subsequent (deep)fake news exposures.
• Manipulation of news modality, presenting celebrity (deep)fake news in either text or video form within a simulated social media environment.
• Measurement of authentication difficulty, self-confidence in truth discernment, and susceptibility to later (deep)fake content.

Findings

• Difficulty during initial authentication carried over to shape responses to subsequent fake content.
• Carryover was modality-congruent: it manifested when initial and subsequent fake news shared the same format, not across formats.
• Experienced difficulty influenced both metacognitive self-confidence in truth-discernment and downstream susceptibility to disinformation.

Connections

This paper sits within a wave of work on synthetic media and deepfake susceptibility, complementing Appel2026-qr and Di-Domenico2026-zq on AI-generated content perception, and Hameleers2026-mc on multimodal disinformation. Its focus on metacognitive confidence in discernment relates to van-der-Linden2026-jt and inoculation-style work, while the sequential-exposure framing offers a complement to single-shot susceptibility studies such as Mosleh2024-op and DeVerna2025-dl.