社交媒体平台虚假信息的泛滥已成为不可忽视的治理问题。近年来，越来越多的社交平台开始借助警示标签辅助虚假信息治理，为内容贴上警示标签也被视为阻止虚假信息传播的有效手段，逐渐成为人们研究和讨论的焦点。

本研究希望探究在中国社交平台环境下，虚假信息警示标签会产生何种认知效果，能否有效降低虚假信息的感知可信度、转发意愿。同时研究还借助启发式-系统式思维模型（HSM）来更好地理解平台用户和警示标签的互动过程。值得注意的是，目前我国社交平台警示标签的覆盖率相对较低，很大一部分虚假信息暂未被标记警示，因此本研究也探究了警示标签对无标签信息的认知影响，来综合分析警示标签的实际效果。为实现本次研究目标，本研究以513名中国社交平台用户为样本，通过2（内容判定：争议/虚假）×2（呈现形式：嵌入/覆盖）的组间在线实验，考察四种警示标签组合对感知可信度、转发意愿以及思维水平的影响。

研究结果表明，警示标签显著降低了虚假信息的感知可信度，其中虚假判定×嵌入式标签组合效果最好。标签并不能降低附标签虚假信息的转发意愿。本次研究并未观察到警示标签的“副作用”，标签的存在同样能够降低无标签虚假信息的感知可信度和转发意愿，从而更好地辅助虚假信息治理，但警示标签也会降低无标签真实信息的转发意愿。此外，警示标签能有效激活系统式思维水平，抑制启发式思维水平，让被试面对虚假信息时更加仔细、谨慎。中介效果检验显示，系统式思维能正向预测虚假信息感知可信度，启发式思维能负向预测感知可信度，标签会通过影响思维水平来降低虚假信息的感知可信度，因此HSM理论能帮助我们更好地理解警示标签的作用机制。本研究初步证实了中国社交媒体环境下，虚假信息警示标签的认知效果，丰富了虚假信息治理实践研究，有助于平台不断探寻标签治理的最佳实践方式。

The proliferation of false information on social media platforms has become a governance problem that cannot be ignored. In recent years, more and more social platforms have begun to assist the governance of false information with the help of warning labels, and labeling content with warning labels is also regarded as an effective means to stop the spread of false information, which has gradually become the focus of research and discussion.

In this study, we hope to investigate what kind of cognitive effects false information warning labels have in the Chinese social platform environment, and whether they can effectively reduce the perceived credibility and forwarding intention of false information. The study also utilizes the Heuristic-Systematic Thinking Model (HSM) to better understand the interaction process between platform users and warning labels. It is worth noting that the coverage rate of warning labels on social platforms in China is relatively low, and a large portion of false information has not been labeled with warnings for the time being; therefore, this study also explores the cognitive impact of warning labels on unlabeled information to comprehensively analyze the actual effect of warning labels. To achieve the goal of this study, this study conducted an online experiment to examine the effects of four warning label combinations on perceived credibility, retweeting willingness, and level of thinking through a 2 (controversy/falsehood) × 2 (embedding/coverage) between-subjects design with a sample of 513 Chinese social media platform users.

The results of the study showed that warning labels significantly reduced the perceived credibility of false information, with the false judgment × embedded label combination having the best effect. The labels did not reduce the willingness to forward the attached labeled false information. In this study, we did not observe the "side effects" of warning labels. The presence of labels can also reduce the perceived credibility and forwarding willingness of unlabeled false information, thus better assisting the governance of false information, but warning labels also reduce the willingness to forward unlabeled true information. In addition, warning labels can effectively activate the level of systematic thinking and inhibit the level of heuristic thinking, making subjects more careful and cautious when facing false information. The mediation effect test showed that systematic thinking positively predicts the perceived credibility of false information, and heuristic thinking negatively predicts the perceived credibility, and labels reduce the perceived credibility of false information by affecting the level of thinking, so the HSM theory can help us better understand the mechanism of warning labels' action. This study preliminarily confirms the perceived effects of disinformation warning labels in the Chinese social media environment, enriches the research on disinformation governance practices, and helps platforms to continuously explore the best practices of label governance.