CWCheng-Jun Wang

Cheng-Jun Wang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Cheng-Jun Wang has a total of 49 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Weaving Behavioral Cocoons in Modern Life (2003–2020): Understanding the Acceleration of Pace of Life and Media Use with Word Embeddings

    byĀ Zhuo Chen, Shengchun Huang, Cheng-Jun Wang

    Modern society is characterized by pervasive feelings of time pressure and a sense of compression in daily activities (Wajcman, 2008). This phenomenon, termed ā€œacceleration of the pace of lifeā€ (Ro...

  • Shut Your Eyes in the Dark! Differential Effects of Pre-Bedtime and Pre-Shuteye Media Use on Sleep Delay and Anxiety

    byĀ Xue-Fei Yan, Zhihao Ma, Yaotian Zhang, Yi-Xin Zhou, Cheng-Jun Wang

    The prevalence of media use and sleep deprivation has raised growing public health concerns (Matricciani et al., 2017). Evidence shows a steady decline in average sleep duration across many countri...

  • Open Access

    Framing the world: Strategic narratives of foreign states in China’s mainstream media (1950–2019)

    byĀ Zhicong Chen, Zhengyi Liang, Zhenyu Wang, Xinya Jiang, Cheng-Jun Wang

    This study investigates how mainstream Chinese media have portrayed foreign states over the past seven decades. While scholars have extensively studied foreign media portrayals of China, little systematic or longitudinal evidence exists on how China’s official party press frames the world across different countries and time periods. Using word embedding techniques, this study traces media representations of major foreign countries (N = 94) on People’s Daily (1950–2019). Results show that media favorability trends in the People’s Daily closely mirror major turning points in Chinese foreign policy. Regression analyses reveal that a country’s economic development, industrialization, and military capacity are major predictors of its media image, while the influence of economic factors shifted significantly over time. These patterns reflect broader geopolitical transformations and China’s evolving global orientation. The findings illuminate how computational approaches can recover the structural and strategic logics of state media framing, with implications for ongoing debates on narrative power, cultural influence, and China’s evolving role in the global information order.

  • Open Access

    TRGCN: a hybrid framework for social network rumor detection

    byĀ Yanqin Yan, Suiyu Zhang, Dingguo Yu, Yijie Zhou, Cheng-Jun Wang, Ke-Ke Shang

    Accurate and efficient rumor detection is critical for information governance, particularly in the context of the rapid spread of misinformation on social networks. Traditional rumor detection relied primarily on manual analysis. With the continuous advancement of technology, machine learning and deep learning approaches for rumor identification have gradually emerged and gained prominence. However, previous approaches often struggle to simultaneously capture both the sequential and the global structural relationships among topological nodes within a social network. To tackle this issue, we introduce a hybrid model for detecting rumors that integrates a Graph Convolutional Network (GCN) with a Transformer architecture, aiming to leverage the complementary strengths of structural and semantic feature extraction. Positional encoding helps preserve the sequential order of these nodes within the propagation structure. The use of Multi-head attention mechanisms enables the model to capture features across diverse representational subspaces, thereby enhancing both the richness and depth of text comprehension. This integration allows the framework to concurrently identify the key propagation network of rumors, the textual content, the long-range dependencies, and the sequence among propagation nodes. Experimental evaluations on publicly available datasets, including Twitter 15 and Twitter 16, demonstrate that our proposed fusion model significantly outperforms both standalone models and existing mainstream methods in terms of accuracy. These results validate the effectiveness and superiority of our approach for the rumor detection task.


  • Open Access

    Is National Identity Harmless to Ingroup Members? Leveraging Ingroup Favoritism and Moral Foundation to Promote Online Fandom Donation

    byĀ Yaotian Zhang, Yixiao Sun, Lihan Yan, Xue-Fei Yan, Cheng-Jun Wang

    Social media is not only a contested space for fandoms vying for attention and influence, but also an important platform for carrying out a range of philanthropic actions. However, existing research overlooks the effects of fans’ ingroup favoritism and outgroup hostility—do these two sides of social identity stimulate or inhibit fans’ prosocial behavior towards ingroup members? Independent and bounded theoretical perspectives on ingroup favoritism and outgroup hostility offer opposing propositions. Moreover, existing research has yet to address how the moral language and frames invoked by fans influence their prosocial behavior. In this study, we propose the social identity theory of moral effects (SIME), which posits that moral foundations play a critical mediating role between social identity and prosocial action. By constructing a four-wave panel dataset of 1,957 Chinese fans, we examine the influence of a broader form of social identity (i.e., national identity) and moral language on fandom philanthropy during public crises to address the theoretical question. We find that ingroup favoritism and individualizing moral language stimulate fandom donations, whereas outgroup hostility and binding moral foundations weaken them. Furthermore, the causal mediation analysis reveals that language based on these moral foundations is a key mechanism, indirectly mediating the effects of social identity on fandom donations. We discuss the theoretical and practical contributions of these findings.

  • Open Access

    TRGCN: A Hybrid Framework for Social Network Rumor Detection

    byĀ Yanqin Yan, Suiyu Zhang, Dingguo Yu, Yijie Zhou, Cheng-Jun Wang, Ke-Ke Shang

    Accurate and efficient rumor detection is critical for information governance, particularly in the context of the rapid spread of misinformation on social networks. Traditional rumor detection relied primarily on manual analysis. With the continuous advancement of technology, machine learning and deep learning approaches for rumor identification have gradually emerged and gained prominence. However, previous approaches often struggle to simultaneously capture both the sequential and the global structural relationships among topological nodes within a social network. To tackle this issue, we introduce a hybrid model for detecting rumors that integrates a Graph Convolutional Network (GCN) with a Transformer architecture, aiming to leverage the complementary strengths of structural and semantic feature extraction. Positional encoding helps preserve the sequential order of these nodes within the propagation structure. The use of Multi-head attention mechanisms enables the model to capture features across diverse representational subspaces, thereby enhancing both the richness and depth of text comprehension. This integration allows the framework to concurrently identify the key propagation network of rumors, the textual content, the long-range dependencies, and the sequence among propagation nodes. Experimental evaluations on publicly available datasets, including Twitter 15 and Twitter 16, demonstrate that our proposed fusion model significantly outperforms both standalone models and existing mainstream methods in terms of accuracy. These results validate the effectiveness and superiority of our approach for the rumor detection task.

  • Open Access

    Behind the Roles We Play: How U.S. Politicians’ Social Media Narrative Identity Relates to Their Dissent Voting Behavior

    byĀ Yixiao Sun, David Matthew Markowitz, Yaotian Zhang, Cheng-Jun Wang

    Inspired by the theoretical tension between Goffman’s frontstage self-presentation and McAdams’s narrative identity theory, this study investigated if politicians’ public social media personas are strategic performances or reflect an internalized identity predicting backstage political behavior. We examined this authenticity paradox via U.S. legislators’ dissent voting, linking the language in 3.26 million tweets from 680 U.S. House Representatives (115th–118th Congresses) to 1.32 million roll-call votes. We used different natural language processing and network analysis techniques to measure narrative (agency, communion) and social identities (partisan interaction, structural holes) with fixed-effects panel models to observe that frontstage narratives predict backstage actions. Both agency and communion on Twitter were significantly associated with more dissent voting. Social identity, particularly higher inter-party interaction and spanning inter-party structural holes (lower Burt’s constraint), also positively predicted dissent. Mediation analysis confirmed social identity factors as key mechanisms linking narrative identity to dissent behavior. These findings suggest politicians’ self-narratives are public narratives that help to internalize an identity, shaping a self-concept that drives identity-congruent political action.

  • Who weaved my behavior cocoon? The impact of digital media use on daily behaviors in an accelerated society

    byĀ Xue-Fei Yan, Zhicong Chen, Linyan Lu, Cheng-Jun Wang

    Whether information communication technologies extend or constrain our daily activities remains an important yet unsettled question. To reconcile the theoretical tensions between displacement and enhancement, this study examines how digital media have shaped the behavior cocoon, defined as a constrained behavioral space where individuals’ daily behaviors become increasingly homogenized (less diverse) and centralized around work. Using data from the United Kingdom Time Use Survey 2016–2020, we find that digital media use for work is likely to displace other activities and increase the formation of behavior cocoons, while digital media use for leisure tends to enhance other activities and inhibit the behavior cocoon. Interestingly, as work-life balance increases, the effect of digital media use for work on the behavior cocoon shifts from intensifying to mitigating. Our work sheds light on the tensions associated with the impact of digital media on modern life, providing new insights into the paradox of social acceleration, and reveals that work-life balance serves as a crucial mechanism for moderating technological tension.


  • Open Access

    The Unintended Consequences of Emancipation: How the Abolition of Slavery Increased Racist Narratives in the American South (1800–1920)

    byĀ Zhenfeng Zhou, Cheng-Jun Wang

    Does the abolition of racial hierarchy undermine or enhance racism? From the contact hypothesis and theory of racial assimilation to structural racism and racial threat theories, these frameworks provide distinct perspectives on the relation between the abolition of racial hierarchy and racism. To address this theoretical puzzle, we examine the impact of the abolition of slavery following the Civil War—a profound disruption to the racial hierarchy—on the development of racism in the American South. Using an original dataset covering 1403 southern counties over a span of 121 years, we find that the abolition of slavery increases the prevalence of racist narratives in the South. This effect is primarily driven by economic competition between freed Black workers and White workers in the labor market. Furthermore, we explore the long-term effects of slavery abolition and demonstrate that even 150 years later, slavery continues to shape modern racial biases.

  • Let Us Not Wallow in the Valley of Despair: The Role of Emotion, Panic, and Sympathy Discourses in Promoting Productive Actions

    byĀ Yaotian Zhang, Yixiao Sun, Gongjing Lu, Zhicong Chen, Cheng-Jun Wang

    Facing severe public crises, individuals might fall into a valley of despair as they navigate overwhelming positive and negative emotions. The Communication Theory of Resilience (CTR) highlights the significance of legitimizing negative emotions and highlighting positive emotions in fostering resilience, while the Stress and Coping Framework of Resilience (SCFR) highlights the negative emotions in cultivating coping actions. However, the emotion dynamics remain underexplored. In this study, we propose the Emotion Dynamics Model of Resilience (EDMR) and employ four-wave panel data in the early stages of the COVID-19 pandemic on social media to test the hypotheses. The results show a U-shaped relationship between negative emotions and online donations, contrasted with an inverted U-shaped relationship between positive emotions and online donations. Further, panic and sympathy discourses amplify the effect of emotions on productive actions. EDMR advances extant understanding of resilience processes by examining the nonlinear effects of emotions and the interaction between emotions and discourse.