At the start of the COVID-19 pandemic, there was no vaccine to cure or slow its impact due to the novelty of the virus, nor were there were any other standardized measures to handle its spread. Yet, despite the detrimental consequences of the pandemic and its impact on people’s lives, the behavior of individuals to combat the pandemic was not necessarily consistent with official guidelines. To make things worse, the pandemic was highly politicized in countries such as the U.S. With a help of a national survey from the U.S., we examine the associations between media literacy variables and willingness to perform recommended COVID-19 related health behavior. Moreover, we also examine the moderating role of conservative media use in this relationship. Our findings show that conservative media use was negatively associated with these protective behaviors, and that both media literacy variables were positively related with willingness to perform recommended COVID-19 related health behavior. Our results show that media literacy can mitigate some of the impact of conservative media use on individuals. Our findings help understand the complexity of protective behavior against the virus during a highly politicized pandemic.

The 2022 Beijing Winter Olympics took place when American public opinion about China was already predominantly negative as media reports had a lot of highlights of human rights violations happening in China. Besides, earlier, the global image of China was undermined by the COVID-19 pandemic. In this study, we explored whether American public opinion about the Olympics and China can be influenced by images of the Olympics shared in the media. Findings from the 2 × 2 between-subject experimental design suggest that the opinions about the Olympics can be predicted by people's beliefs about China's role in the global pandemic. In addition to that, our findings suggest that pre-existing opinions and feelings held by most of the population can mitigate the effects of visual primes.

The increase in verbal and physical aggressions directed at Asian Americans during the COVID-19 pandemic and after has brought into focus a need to understand the reasons for such aggressions and identify interventions. Our study addresses these needs by investigating whether photographs accompanying a news report depicting Asian Americans positively or negatively in the context of the COVID-19 pandemic activate positive or negative perceptions of Asian Americans and whether the valence of the photographs differentially influences intent to intervene in anti-Asian aggressions. In a 2 (image: positive vs. negative) X 2 (political ideology: liberal vs. conservative) mixed between-subject online experiment with adult white Americans, we found that liberals showed more positive perceptions and greater intent to intervene with the negative photos compared to the positive photos, while conservatives showed more negative perceptions and lower intent to intervene with the negative photos. Implications of these findings for activation and activation control theories of primes and biases are discussed. Some recommendations on how the results might guide media content providers to mitigate the spread of anti-Asian aggressions are also provided.

Emotion detection in social media remains challenging, particularly in polarized public debates where emotional expression is often imbalanced across categories. Addressing this challenge, this study compares the performance of multiple computational approaches using a gold-standard dataset of tweets about an ongoing geopolitical conflict. The dataset reflects the authentic, skewed distribution of emotions observed in real-world online discourse. We evaluated lexicon-based methods, classical machine-learning classifiers, deep-learning architectures, transformer models in both fine-tuned and zero-shot configurations, and a zero-shot large language model to assess their effectiveness in capturing both frequent and less frequently expressed emotions. Across approaches, transformer models, especially those fine-tuned for contextual emotion recognition, demonstrated the strongest overall performance, with emotion-specific fine-tuning offering a particular advantage for detecting rare emotion categories. These findings emphasize the importance of evaluating emotion detection methods under realistic class imbalance and highlight both the comparative strengths and limitations of widely used modeling strategies in applied social media research. This study advances emotion analysis and computational social science by offering practical guidance for selecting appropriate emotion detection methods in complex, imbalanced social media contexts.