Social evaluation is fundamental to everyday interactions, yet our understanding has been constrained by fragmented theories and the lack of a scalable method for tracking group attitudes in real time. This paper resolves this methodological gap by introducing and validating a computational framework that empirically synthesizes three major theoretical models (Stereotype Content Model, Dual Perspective Model, and Semantic Differential) within a unified word embedding space. We demonstrate that social evaluation is structured by two core latent dimensions: Warmth-Communion-Evaluation (WCE), capturing affective and moral judgments, and Competence-Agency (CA), reflecting perceptions of ability and effectiveness. To validate its real-world utility, we apply this framework to U.S.-based Twitter posts about Chinese and Japanese individuals before and during the COVID-19 pandemic. Our analysis reveals that while perceptions of competence (CA) remained stable, affective evaluations (WCE) of Chinese individuals declined sharply, a dynamic not observed for Japanese individuals. This work offers a robust, scalable instrument for tracking intergroup attitudes during crises and provides a crucial bridge between social psychological theory and computational social science, enabling the real-time analysis of intergroup dynamics.

This study analyzes 15,000 user comments from ideologically distinct news platforms—Fox News (right-leaning), ABC News (left-leaning), and BBC News (centrist)—to investigate how political ideology shapes emotional expression and engagement following the July 2024 assassination attempt on President Donald Trump in Pennsylvania. We examine whether right-leaning comments exhibit elevated anger and hate speech, whether left-leaning comments show greater empathy and positive affect, and how emotional intensity influences engagement metrics such as likes and shares. Using computational science methods, we trace the evolution of emotional tone and toxicity over time, focusing on affective polarization triggered by a high-salience violent event. Our findings reveal that ideological alignment amplifies engagement with congruent emotional content, reinforcing echo chambers and intensifying online hostility. We address gaps in understanding how ideological echo chambers amplify polarized emotions and online hostility following a high-profile violent event, emphasizing the place of emotions in contemporary political communication and digital news consumption.

This study investigates which members of the 118th U.S. Congress adopt and use Threads and TikTok, and what political, demographic, and constituency-level characteristics explain this variation. Grounded in diffusion of innovation theory, we ask: What factors predict the adoption and use of these emerging platforms? We compiled original data on all members of Congress ( N = 535) by collecting social media account information from official congressional websites and manually verifying platform presence. Adoption was measured as a binary variable, and usage as the number of posts made through November 2023. Using probit and OLS regression models, we tested predictors including party affiliation, age, race, leadership status, and prior digital engagement. The empirical analyses reveal that Democrats and younger legislators are more likely to adopt Threads and TikTok. Prior digital engagement consistently predicts usage on both platforms. Notably, racial identity plays a critical role: non-white members are more likely to adopt and use TikTok, while white members are more likely to use Threads. This study offers the first empirical analysis of congressional adoption and usage of Threads and TikTok. Our findings demonstrate that platform choice is shaped by identity, institutional context, and political strategy. These findings offer new insights into the determinants of early platform adoption among U.S. congress members and the importance of aligning communication choices with constituent behavior and platform culture.

Digital platforms now serve as crucial archives for analysing societal trends, yet their research APIs pose methodological challenges. This study critically evaluates TikTok Research API through comparative analysis of 6,373 videos from 14 creators in the United States and United Kingdom (2020–2022), contrasting API-derived outputs with manual collection and third-party analytics. The API demonstrated scalability, retrieving more videos than alternative methods and providing 22 variables, including eight unavailable elsewhere. However, limitations were substantial: transcriptions covered about 10% of the content, with more transcripts returned from American male creators. Engagement metrics exhibited inconsistent accuracy across data sources, with the API showing systematically lower view counts but higher comment and share counts compared to manual collection. The number of videos varied depending on sample composition, indicating that small changes in inclusion criteria could shift outcomes disproportionately. These results highlight systematic inconsistencies, showing why multi-method approaches remain necessary despite automation. While TikTok Research API offers valuable scale and ethical compliance, its demographic biases and metadata inconsistencies compromise validity. The study advocates integrated auditing protocols and targeted API refinements to improve representativeness and accuracy in platform-based research.

This study investigates the impact of self-efficacy among marketing professionals with digital tools affecting their strategic flexibility (ability to adjust marketing strategies to changing trends) and career adaptability (willingness to learn or change roles), and how these variables influence the adoption of metaverse marketing tools. Data were gathered using a cross-sectional online survey of 395 marketing and digital strategy professionals working in the e-commerce, digital marketing, information technology, and educational-technology sectors in China. Structural Equation Modeling (SEM) was used to examine a moderated mediation model between self-efficacy, strategic flexibility, career adaptability, and metaverse marketing adoption. Results show that Marketing self-efficacy significantly predicts strategic flexibility (β = 0.35, p < 0.001) and career adaptability (β = 0.30, p < 0.001) which fully mediate its impact on adoption, indicating an indirect behavioral path. Interestingly, avatar anthropomorphism strengthens these relationships (β = 0.20, p < 0.01 for strategic flexibility; β = 0.18, p < 0.01 for career adaptability) with higher human-likeness intensifying the mediated effects and highlighting the impact of anthropomorphic cues in virtual environments. This study provides insights for equipping professionals for digital transformation and offering actionable strategies for designing avatars to improve technology adoption in virtual marketing environments. The findings highlight that well calibrating avatar features—responsiveness, realism, and feedback—can enhance perceived trust and usability and social presence. Overall, this study extends research on virtual environments by identifying avatar anthropomorphism as a key boundary condition in technology adoption and offering design implications for AI-driven metaverse interfaces.

Objective: This paper analyzes gender equality across countries in the year 2024 by using the GGGI, with the intention of disentangling the unseen structural and non-deterministic patterns. Instead of repeating the process of calculating the index, it is openly recognizing the compositional feature of the GGGI and the unseen similarities between the indices. Methods: This research employs a global cross-sectional study of 146 countries over the four primary GGGI sectors: economic participation, education, health and survival, and empowerment. Where OLS is only employed as a diagnostic test, as its almost perfect fit (R 2 ∼1) is squarely mechanical and lacks relevance for inference. Apart from ensemble models employed for predictions, K-means clustering, SHAP analysis, and GridSearchCV optimization are also used. Findings: The out-of-sample predictions demonstrate high levels of predictive accuracy, with Gradient Boosting models yielding an R 2 of approximately 0.90 and an RMSE of approximately 0.045, indicating that there is significant nonlinear information beyond index aggregation. Unsupervised clustering techniques show that there are seven distinct country clusters that go beyond traditional geographic and income divisions, which can be identified with more than 93% accuracy. The SHAP results show that empowerment and economic participation are drivers, while there is insignificant variation in healthcare. Contribution: This study identifies the boundaries of regression analysis in index research, as well as the advantages of machine learning analysis in determining structural patterns related to gender equity.

This paper describes a three-part framework to study how language technologies elucidate and shape linguistic relations in society. Reframing a mountain of evidence about language bias in LLMs, we introduce the concept of linguistic affordances to attend to how an object can shape social relations through language. First, we contextualize how language ideologies inform social relations in a particular setting. Next, we examine how language ideologies shape the construction of the linguistic affordances of a language technology. Finally, we examine how the linguistic affordances of language technologies lead to new associations that link language and social worth. We describe how this framework can inform both the study of language technologies and the use of language technologies in social science. We demonstrate the framework with two examples: the use of LLMs in college admissions and the adoption of LLMs in scientific publishing.

Previous research has highlighted that encounters with deepfakes induce uncertainty, skepticism, and mistrust among audiences. In this study, we relate perceived deepfake exposure to media cynicism. Deepfakes shake users’ sense of reality, increasing a need to rely on epistemic authorities, such as journalistic media, while raising fears of manipulation. Based on uncertainty management theory, we propose that two “epistemic virtues” moderate the relationship between deepfake exposure and media cynicism: self-efficacy and intellectual humility. In a survey of 1421 German internet users, we find that perceived deepfake exposure positively relates to media cynicism. Intellectual humility does not dampen this relationship. Deepfake detection self-efficacy may be more harmful than helpful in preventing media cynicism. We discuss these findings in the context of research indicating that users tend to overestimate their ability to detect deepfakes and the challenges the novel deepfake technology poses to audience trust in a digital information ecosystem.

This research note addresses a methodological gap in the study of large language models (LLMs) in social sciences: the absence of standardized data extraction procedures. While existing research has examined biases and the reliability of LLM-generated content, the establishment of transparent extraction protocols necessarily precedes substantive analysis. The paper introduces a replicable procedural framework for extracting structured political data from LLMs via API, designed to enhance transparency, accessibility, and reproducibility. Canadian federal and Quebec provincial politicians serve as an illustrative case to demonstrate the extraction methodology, encompassing prompt engineering, output processing, and error handling mechanisms. The procedure facilitates systematic data collection across multiple LLM versions, enabling inter-model comparisons while addressing extraction challenges such as response variability and malformed outputs. The contribution is primarily methodological—providing researchers with a foundational extraction protocol adaptable to diverse research contexts. This standardized approach constitutes an essential preliminary step for subsequent evaluation of LLM-generated content, establishing procedural clarity in this methodologically developing research domain.

Information disorder (IDO) presents a persistent challenge to society, necessitating innovative approaches to understanding its dynamics beyond just merely detecting it. This study introduces a theory-driven framework that integrates advanced natural language processing (NLP) with deep learning, utilizing the target-based emotion–stance analysis (TESA) approach to analyze emotion and stance dynamics within IDO content. Complementing TESA, interactive graph generation (IGG) is applied for scalable and interpretable qualitative analyses. Employing a mixed-methods approach, the study leverages TESA for target-centric emotion and stance analysis, evaluating target-based classifiers on both human-annotated and synthetic datasets. Additionally, the study explores synthetic data generation using generative AI to enrich the analysis, applying IGG to map complex data interactions. The study also found that integrating synthetic data developed from human annotations enhanced model performance, particularly for emotion classification tasks. Results demonstrate that IDO narratives significantly differ from non-IDO narratives, frequently leveraging negative emotions such as anger and disgust to manipulate public perception. TESA proved effective in capturing these nuanced variations, while IGG facilitated the triangulation of such findings via the scalable interpretation of emotional narratives, revealing that IDO content often amplifies polarizing and antagonistic perspectives. By combining TESA and IGG, this research emphasizes the importance of using NLP to extract and examine the emotional and stance nuances toward targets of interest within IDO context. This approach not only deepens theoretical insights into IDO’s persuasive mechanisms but also supports the development of practical tools for analyzing and managing the influence of IDO on public discourse.