Platformed Publics and Image Transformation: Factor Analysis of Foreign YouTubers’ China Discourse (2015–2025)

Amid global polycrises—public health emergencies, energy shocks, geopolitical tensions, and climate degradation—algorithm-driven video platforms have evolved into grassroots public spheres separate from elite mainstream media. YouTube, as the largest global user-generated video platform, accumulates massive long-term discourse produced by independent foreign vloggers, whose unscripted narratives offer perceptions of China that diverge from state-led official media portrayals. Drawing upon three core theoretical lenses—platform public theory, Wodak’s discourse-historical approach (DHA), and discursive news value framework—this study establishes a ten-year longitudinal corpus covering foreign YouTubers’ full video transcripts published between 2015 and 2025. After manual thematic coding and affective labelling of textual materials, this research transforms qualitative discourse indicators into standardized measurable variables and adopts exploratory factor analysis (EFA) alongside confirmatory factor analysis (CFA) via SPSS 28 to extract latent dimensional factors of overseas creators’ cognition of China’s national image. Diachronic comparative analysis further traces the shifting weights, emotional polarities and thematic priorities of each factor across three chronological phases. The results reveal six core latent dimensions that structure foreign vloggers’ overall perceptions: economic development perception, technological innovation evaluation, ecological and climate governance judgment, cultural identity recognition, social livelihood observation, and international diplomatic cognition. From 2015 to 2025, the dominant discourse factors exhibited clear stage-based shifts: early narratives centred on China’s economic miracle, mid-stage discourses tilted toward geopolitical and institutional conflict frames under polycrisis impacts, while late-stage texts gradually highlighted China’s renewable energy and climate governance practices. Platform algorithms, cascading global crises, and cross-cultural cognitive biases jointly account for these perceptual shifts over the decade. This study integrates corpus-assisted critical discourse analysis and multivariate statistical factor methods, extends the application scope of platform public and discourse-historical theories to long-form video vlog research, and delivers empirical evidence for understanding grassroots transnational image construction in the algorithmic platform era.