Accuracy Prompts and Health Misinformation: Differential Effects by Dimension and Platform

Non-evidence-based health content circulates widely on TikTok, YouTube, and Instagram, posing threats to public health. Although accuracy prompts consistently reduce misinformation sharing in political news contexts, their effects on misleading health content remain underexplored in Latin American Spanish-speaking populations. A between-groups experimental design with random assignment was implemented with 300 university students (n = 150 control, n = 150 experimental), who evaluated nine social media posts before and after the experimental group received a brief accuracy prompt. Perceived trust, sharing resistance, and perceived veracity were measured (α = .770 to .859); a confirmatory factor analysis indicated that trust and veracity were not empirically distinguishable and were combined into a single Epistemic Evaluation dimension. The accuracy prompt significantly reduced Epistemic Evaluation scores, indicating lower perceived trust and veracity toward misleading health content (d = 0.832, p < .001), and produced a smaller significant increase in Sharing Resistance (d = 0.346, p = .003). The effect on Epistemic Evaluation was consistent across platforms, whereas a significant Condition × Platform interaction emerged for Sharing Resistance, significant only on Instagram; because platform identity was partially confounded with narrative content, this pattern is interpreted as preliminary. Implications for theory, education, and future research are discussed.