SHENGCHUN HUANG

My research examines how AI and algorithmic systems reshape political information environments. I am particularly interested in how systems of curation structure what becomes visible to citizens, how people perceive those environments, and what these changes mean for political learning and judgment.

My current work focuses on two related questions:

(1) how algorithmic curation structures opportunities to encounter civic information, and
(2) how people perceive and navigate increasingly personalized and AI-mediated information environments.

My work combines computational social science with algorithm audits, field and survey experiments, and behavioral data. Across these approaches, I seek to make the otherwise hidden structure of contemporary information environments empirically observable.

Peer-Reviewed Publications

(† indicates equal first authorship; * indicates corresponding author)

Chen, Z.†, Huang, S.†, Wang, C. (in press). Weaving Behavioral Cocoons in Modern Life: Understanding Acceleration of the Pace of Life with Word Embeddings. Mass Communication and Society.

Abstract: The acceleration of the pace of life appears to be a prominent feature of modern society. Research has explored several driving forces behind this phenomenon, one of which is information and communication technologies (ICTs). However, there is a lack of a systematic, quantifiable way to either examine the extent to which the life pace is being accelerated or investigate the relationship between the use of digital technologies and acceleration. This study aims to fill this gap by providing a set of empirical indicators of acceleration and testing their associations with ICTs. Using the datasets of the American Time Use Survey (ATUS) from 2003 to 2020, this study measures acceleration from a computational approach. In particular, this study develops a novel measure—behavioral cocoon— with word-embedding to capture the degree of “time-space compression”. This research finds that the extent of acceleration differs across different sociodemographic groups. And the use of ICTs demonstrates nuanced relationships with it. In general, either instrumental or recreational use of ICTs is associated with a more swift and mobile way of life, implying a liberating power against the modern time pressure. Yet, the role of ICTs is limited by other factors such as the living expenditure of the year.

Wang, S.†, Huang, S.†, Zhou, A., & Metaxa, D. (2024). Lower Quantity, Higher Quality: Auditing News Content and User Perceptions on Twitter/X Algorithmic versus Chronological Timelines. Proc. ACM Hum.-Comput. Interact., 8(CSCW2), 507:1-507:25. https://doi.org/10.1145/3687046

Abstract: Social media personalization algorithms increasingly influence the flow of civic information through society, resulting in concerns about "filter bubbles", "echo chambers", and other ways they might exacerbate ideological segregation and fan the spread of polarizing content. To address these concerns, we designed and conducted a sociotechnical audit (STA) to investigate how Twitter/X's timeline algorithm affects news curation while also tracking how user perceptions change in response. We deployed a custom-built system that, over the course of three weeks, passively tracked all tweets loaded in users' browsers in the first week, then in the second week enacted an intervention to users' Twitter/X homepage to restrict their view to only the algorithmic or chronological timeline (randomized). We flipped this condition for each user in the third week. We ran our audit in late 2023, collecting user-centered metrics (self-reported survey measures) and platform-centered metrics (views, clicks, likes) for 243 users, along with over 800,000 tweets. Using the STA framework, our results are two-fold: (1) Our algorithm audit finds that Twitter/X's algorithmic timeline resulted in a lower quantity but higher quality of news -- less ideologically congruent, less extreme, and slightly more reliable -- compared to the chronological timeline. (2) Our user audit suggests that although our timeline intervention had significant effects on users' behaviors, it had little impact on their overall perceptions of the platform. Our paper discusses these findings and their broader implications in the context of algorithmic news curation, user-centric audits, and avenues for independent social science research.

Huang, S.* & Yang, T. (2024). Auditing Entertainment Traps on YouTube: How Do Recommendation Algorithms Pull Users Away from News. Political Communication, 41(6), 903–920. https://doi.org/10.1080/10584609.2024.2343769

Abstract: Recommendation algorithms that customize information feeds for individuals have raised concerns about exacerbating inequalities in news exposure among citizens. In response to these concerns, we conducted an audit study on YouTube to analyze the algorithmic impact on curating news versus other content topics. We examined over 1.7 million YouTube video recommendations audited in 2019 and developed novel analysis approaches including network analysis and Markov chains. Results show that recommendation algorithms may potentially redirect users away from news content through two influence pathways: (1) the “topical filter bubbles,” wherein entertainment content has a higher probability of being recommended over news content in a self-reinforcing manner; and (2) “algorithmic redirection,” wherein the probability of entertainment videos being recommended after a news video is much higher than that for the opposite. Overall, YouTube recommendation algorithms have a higher probability of recommending entertainment videos than news. The findings imply essential biases in algorithmic recommendations on digital platforms beyond amplifying users’ preferences.

Huang, S. *† & Yang, T.† (2024). No Trade-offs between News and Entertainment: Evidence from Online Engagement Data. New Media & Society, 26(2), 899–920. https://doi.org/10.1177/14614448211063899

Abstract: In today’s high-choice media environment, some scholars are concerned that people selectively consume media content based on personal interests and avoid others, which might lead to audience fragmentation across different content genres. Individually, there might be trade-offs between those genres, especially entertainment versus news. This study analyzed a large user engagement dataset (~40,000 users’ comments) collected from the Chinese information application Toutiao, one of the most popular information distribution platforms in China. The results showed that (1) the commenters were not fragmented between content genres, and (2) the users’ news engagement was positively associated with their entertainment engagement. The findings indicate that the availability of high media choices will not reduce the news engagement of those who have strong interest in entertainment. Instead, news engagement might increase alongside the augmentation of the sum of information engagement. Finally, we discussed the differences between relative news engagement and absolute news engagement.

Zhang, Y., Huang, S.*, & Li, T. (2023). ‘Push-and-pull’ for Visibility: How do Fans as Users Negotiate over Algorithms with Chinese Digital Platforms? Information, Communication & Society, 26(2), 321–339. https://doi.org/10.1080/1369118X.2022.2161829

Abstract: In the algorithmic era, both users and the platform battle for visibility. Chinese fans are savvy users who explore the hidden algorithms behind platform functions. With the collectively developed algorithmic imaginary, digital fandom communities negotiate with the platform over algorithms to optimize the visibility of celebrities they endorse. Drawing from participatory observation and semi-structured interviews in Chinese online fandoms of an idol group, INTO1, we detailed how fans as digital users collectively explore, interpret, and creatively utilize algorithms to increase their idol’s visibility. We conclude that visibility, as a representation of algorithm power, is co-defined through the constant push-and-pull between digital users and the platform. This paper contributes to both algorithm and fandom studies by describing large-scale non-professional users’ daily construction of the algorithmic imaginary in the unique context of Chinese fandom and beyond. It also discusses broader civic implications of fans’ algorithmic practices to wider digital users in China.

Ahn, C., Huang, S. & Lelkes, Y. (2026). Strategic Animosity: How Social Context Shapes Expressions of Political Hostility. The Economic Journal. https://doi.org/10.1093/ej/ueag055

Abstract: Why do partisans share vitriolic information? Scholars have pointed to several possible explanations, including social identity signaling, prevailing social norms, and an evolved tendency to consume and prefer negative news. Still, no study simultaneously evaluates these three explanations or examines interpersonal dynamics. We conducted a large-scale interactive online experiment that paired individuals with other participants and measured what type of news they shared with partners. Two factors were manipulated: the relationship between partners’ political identities and whether this information is disclosed. We find the sharing of negative news is largely motivated by dyadic norms—people are far more likely to send negative information if such information was shared in the previous round. Moreover, social identity theory provides only a partial explanation for sharing behavior—shared partner-sender identity mattered, but not necessarily for negativity. People have a slight overall preference for negativity, which also serves a social function by increasing trust.

Work in Progress

Huang, S., Wang, S., Zhou, A., & Metaxa, D. Visible News Exposure: Exploring Partisan Selectivity in Online Information Ecosystem. [under review]

Ye, Q., Duan, Z., Huang, S. Leveraging Large Language Models in Message Stimuli Generation and Validation for Experimental Research. [under review]

Abstract: Despite the wide application of message stimuli in communication experiments, creating effective stimuli is often challenging and costly. However, the advent of generative artificial intelligence (AI) and large language models (LLMs) suggests great potential to facilitate this process. To advance AI-assisted communication research, we examined the performance of ChatGPT (GPT-4) in generating message stimuli for experimental research. Through four pre-registered experiments, we compared GPT-generated stimuli with human-generated stimuli in (1) manipulating target psychological variables (discrete emotions and moral intuitions) and (2) controlling unintended variables. We found that GPT-generated message stimuli performed equivalently to or even surpassed human-generated stimuli in manipulating target variables, while the performance in controlling unintended variables was mixed. Our study suggests that LLMs can generate effective message stimuli for experimental research. This research serves as a foundational resource for integrating generative AI in stimuli generation across various communication contexts, with its effectiveness, opportunities, and challenges discussed.

Book Chapters & Reviews

Yang, G., Huang, S., Fang, H., Ji, D., Chen, J., & Wang, W. (2025). The Ambivalent Art of Living with Chinese Social Media: Digital Vulnerability and Practices of Self-Care. In Decolonising Approaches to Users and Audiences in the Global South. Routledge.

Abstract: Can subalterns live with care and dignity with social media in non-Western societies? Can Chinese social media users not live in states of oppression? In this chapter, we study a sample (n = 106) of young and educated internet users in metropolitan China to understand how they perceive and manage the benefits and vulnerabilities of social media use. We find that in their mundane digital practices, these internet users practice an art of living with technology. They are more sophisticated – or artistic – in their use of social media than they are given credit for. While recognizing the benefits of social media, they are vulnerable to three main risks: privacy violations, misinformation, and overdependency on social media. They cope with these vulnerabilities with caution and care, exhibiting a sense of reflexivity and agency, as well as of ambivalence. This study contributes empirical and theoretical insights to the broader literature on the ethics and practices of self-care and agency in the age of digital vulnerabilities. The decolonizing thrust in this study lies in its insistence that human actors in non-Western societies, who tend to be given agency only when they are portrayed as dissidents against oppressors, are in fact actively engaged in practices of self-care as practices of freedom.

Huang, S. (2024). Nick Seaver, Computing Taste: Algorithms and the Makers of Music Recommendation. International Journal of Communication, 18, 4–4.

Page updated

Google Sites

Report abuse