30:T480,

Topic models have been widely used by researchers across disciplines to automatically analyze large textual data. However, they often fail to automate content analysis, because the algorithms cannot accurately classify individual sentences into pre-defined topics. Aiming to make topic classification more theoretically grounded and content analysis in general more topic-specific, we have developed Seeded Sequential Latent Dirichlet allocation (LDA), extending the existing LDA algorithm, and implementing it in a widely accessible open-source package. Taking a large corpus of speeches delivered by delegates at the United Nations General Assembly as an example, we explain how our algorithm differs from the original algorithm; why it can classify sentences more accurately; how it accepts pre-defined topics in deductive or semi-deductive analysis; how such ex-ante topic mapping differs from ex-post topic mapping; how it enables topic-specific framing analysis in applied research. We also offer practical guidance on how to determine the optimal number of topics and select seed words for the algorithm.

31:T548,

TikTok—the world’s most downloaded app since 2020, has become a place for more than silly dancing and lip-syncing. TikTok users are increasingly turning to TikTok for news content. Meanwhile, news publishers are embracing TikTok to reach a younger audience. We aim to examine the content strategy adopted by the most-followed news publishers on TikTok and how effective their TikTok strategy is in spurring audience engagement in terms of liking, commenting, and sharing. This study retrieved 101,292 TikTok news videos as of November 22, 2022. With the help of computer vision, natural language processing, and sentiment analysis, we found that TikTok news videos containing negative sentiment and more second-person view shots are associated with significantly higher audience engagement. In addition, this study demonstrated that the TikTok video features and engagement levels differ between the news publishers and other TikTok creators. Moderator analysis shows that both the effect of negative sentiment on engagement and the effect of the second-person view on engagement are moderated by the TikTok account type. The impact of negative sentiment and second-person view on engagement behaviors becomes smaller or even insignificant for news publisher TikTok videos. Theoretical and practical implications are discussed in this study.

32:T4d1,

One of the greatest challenges for public health campaigns is communicating health risks due to the existence of psychological distance. Using COVID-19 as a context, this study designed and tested virtual reality (VR) campaigns based on construal level theory. It assessed the immediate and after-effects of VR on COVID-19 preventive intentions/behaviors and risk perceptions. A total of 120 participants were randomly assigned to see one of four messages: a VR message emphasizing self-interest, a VR message emphasizing other-interest, a print message emphasizing self-interest, or a print message emphasizing other-interest. Preventive intentions/behaviors were assessed at three different times: before, immediately after, and one week after the experimental treatment. Immediately following message exposure, participants exposed to the VR messages perceived a higher level of self-risk than those exposed to print messages. Disgust and fear mediated these effects. One week following message exposure, unvaccinated participants exposed to the VR messages had a higher intention to get vaccinated than those exposed to print messages. Recommendations on how to effectively utilize VR in health interventions are provided.

33:T5ca,