This study investigates the micro-level structural differences in online review affect between social media and e-commerce platforms, exploring the underlying psychological mechanisms that drive consumers to selectively express polarised reviews across these media.
Utilising large language model-driven aspect-based sentiment analysis, this study quantifies an initial dataset of 65,493 smartphone online reviews from TikTok, Bilibili, Jingdong and Suning to identify aspect-level affective differences. To explain the observed differences, Study 2 draws on uses and gratifications theory and conducts a 2 × 2 between-subjects experiment, using wireless headphones as the focal product and TikTok and Taobao as the platforms, to examine how media environments moderate the translation of consumption experiences into review intentions.
Affective differences extend to specific cognitive attributes, with social media platforms exhibiting a significantly more negative and heterogeneous affective landscape, whereas e-commerce platforms demonstrate overwhelming positivity. These differences are driven by a dual-pathway mechanism. Social media platforms primarily activate self-enhancement motivation, which facilitates the translation of negative experiences into negative reviews while weakening the translation of positive experiences into positive reviews. Conversely, e-commerce platforms primarily activate economic motivation, which facilitates the translation of positive experiences into positive reviews while reducing the likelihood of negative experiences translating into negative reviews.
This study extends the investigation of review affect from within-medium comparisons to cross-media analysis. By shifting from macro-level indicators to micro-level attribute evidence, this study reveals a more comprehensive picture of cross-media affective bias. By extending the application of uses and gratifications theory from content consumption to content production, it offers a novel mechanistic explanation for this phenomenon.
