How the crisis of trust in experts occurs on social media in China? Multiple-case analysis based on data mining

The communication of scientific knowledge is crucial, particularly in times of uncertainty. Experts play a vital role in this process, but declining public trust in experts is a growing concern. Social media, while expanding access to information, also amplifies misinformation, thereby further challenging trust in experts. This study focuses on the context of China, a country with centralized leadership and unique characteristics of social media participation. The COVID-19 pandemic offers a rich case study for analyzing the dynamics of trust between the public and experts on social media in this specific environment. While expert advice was crucial during the pandemic, inconsistencies and conflicting information on social media, sometimes fueled by political pressures, contributed to the erosion of public trust. The study aims to analyze public attitudes toward scientific knowledge communicated by experts on Weibo, China's leading social media platform, to gain a deeper understanding of the factors driving this crisis of trust. The study utilizes data mining and sentiment analysis to examine public responses to expert statements, seeking to identify the specific aspects that trigger public skepticism and rejection of expert opinions.

Existing research highlights the importance of public trust in experts for effective science communication. Trust is crucial for simplifying social complexity and reducing uncertainty, but it is easily undermined by conflicting information or perceived unreliability of experts or systems. Studies have shown declining public trust in experts across various fields, such as genetic engineering, climate change, and vaccination. Social media plays a pivotal role, expanding communication channels but also increasing the spread of misinformation. Some studies attribute the crisis of trust to public irrationality or emotional responses, while others emphasize the role of media manipulation in shaping public perceptions. The impact of information sources and characteristics on trust, particularly during the COVID-19 pandemic, has also been widely discussed. While research on trust in expert systems is extensive in democratic contexts, less attention has been given to authoritarian contexts like China, where the government plays a significant role in information control and agenda-setting. This study bridges this research gap by examining the phenomenon within the Chinese context.

This study employs a multiple-case analysis design, examining nine cases of expert-communicated scientific knowledge on Weibo between November 11, 2022, and January 22, 2023. Cases were selected based on three criteria: high relevance to the new coronavirus, high discussion volume (over 5000 comments and influence exceeding half a million), and high visibility (posts by presenters not deleted or comments not blocked). Data collection focused on public comments under posts by the Weibo hashtag presenters, excluding duplicates and irrelevant comments. A total of 36,535 comments were collected. Sentiment classification was performed using a K-Nearest Neighbor (KNN) algorithm trained on a dataset of labeled public comments from other Weibo posts related to the cases. The algorithm categorized comments as either positive or negative, providing a measure of public sentiment towards the experts. Latent Dirichlet Allocation (LDA) topic modeling was then used to cluster the comments into different topics based on their content, identifying dominant themes in the public's response. By combining the sentiment analysis results with the LDA topic model output, the researchers were able to characterize the public’s attitudes towards the expert statements in each case. This combined approach facilitated a comprehensive analysis of public positions and emotional responses to expert-disseminated information. The nine cases varied in their subject matter, including expert statements on the virulence of Omicron, the transition to Class B management of COVID-19, the prediction of future outbreaks, and advice on home remedies.

The analysis revealed several key findings. Firstly, the public's negative reactions to expert statements often stemmed from a perceived lack of rigor, transparency, and data support in the expert's claims. Experts' failure to clearly articulate the limitations of their findings and the uncertainties associated with the evolving pandemic exacerbated the situation. Secondly, the study highlighted the conflicting expert opinions, particularly regarding the virulence of the Omicron variant and the timing and effectiveness of the various policy changes. This inconsistency, perceived by the public as a lack of scientific consensus, further eroded trust. Thirdly, the study revealed the blurring of boundaries between experts and commercial interests or political influence in some cases. The public was particularly skeptical when experts appeared to be promoting products or aligned with specific political agendas, suspecting bias and ulterior motives. Fourthly, the media played a significant role in shaping public perceptions. In several instances, media reports misinterpreted or oversimplified expert statements, potentially misleading the public and contributing to a climate of distrust. This misrepresentation was not always consciously malicious but rather a result of the media's own limited understanding of the scientific nuances or a desire to capture attention through sensationalism. Finally, the study identified irrational emotional outbursts and populist sentiments among certain segments of the public. These expressions of anger, distrust, or outright rejection of expert opinions often lacked a factual basis, reflecting a deeper societal distrust of authority figures and a preference for unsubstantiated alternative narratives.

The study's findings confirm that the crisis of trust in experts on social media is multifaceted, originating from the intertwined actions and interactions of experts, media, and the public. The study sheds light on the importance of transparency, rigorous evidence-based communication, and careful consideration of the public interest by experts. It also highlights the media's crucial role in accurate and unbiased reporting of scientific knowledge. The public's emotional responses and populist tendencies underscore the need for science communication strategies that address these challenges. These findings resonate with existing literature on the politicization of science and the growing influence of 'post-truth' discourse on public discourse. The Chinese context adds a unique layer of complexity due to the centralized government's role in information control and policy adjustments. Understanding how political trust influences public trust in experts is another crucial area for future research.

This study demonstrates that the crisis of trust in experts on Chinese social media is a complex phenomenon driven by the interplay between experts, media, and the public. The key recommendations include a greater emphasis on transparency and evidence-based communication from experts; a commitment to unbiased and accurate reporting by the media; and the development of science communication strategies that engage effectively with the public's emotions and concerns. Future research should focus on longitudinal studies to track the evolution of public trust over time, investigate the impact of different communication strategies on trust, and explore ways to bridge the gap between expert knowledge and public understanding. The limitations of the current study, such as the focus on specific cases and the potential influence of individual characteristics, provide direction for future explorations.

This study has several limitations. Firstly, the focus on comments under posts from Weibo hashtag presenters might not fully capture the range of public opinions. Analyzing comments across all related posts would offer a more comprehensive picture. Secondly, individual characteristics, such as background and political views, were not considered in detail. Further research should incorporate individual-level data to better understand the influence of these factors. Thirdly, the findings might not generalize fully beyond the specific context of China’s COVID-19 policy adjustments, and the study's timeframe could have amplified public sentiment due to the abrupt change in policies. Comparative studies across different national contexts are needed for broader applicability. Finally, this study focuses on external factors affecting public trust; exploring intrinsic factors like pre-existing political trust is crucial for a complete picture.