No abstract

In light of the awareness that we know little about how algorithms are perceived by groups other than those in the mainstream, this study investigates how Chinese gay men on Zhihu generate folk theories of the operation and impact of the platform algorithms. After recruiting 16 long-term users on Zhihu as informants and conducting thematic analysis, two overarching themes are identified: (1) the algorithm as evictor, supported by the users’ folk theories of sidelining, disorganizing, and defaming; and (2) the algorithm as protector, supported by the users’ folk theories of shielding, recognizing, and exclusive networks. Based on the empirical data collected, this study provides inspiration for understanding algorithmic complexity, and challenges the mainstream appeal to break through filter bubbles (information cocoons) by indicating its (hetero)normativity.

The growing interest in combining different approaches to qualitative text and discourse analysis has so far not been met with adapted methodological resources. This article aims to address this gap by developing a methodological framework for combining qualitative text and discourse analysis. First, we introduce four traditions that we identify as four families of methods of text/discourse analysis with different logics: Discourse Analysis, Foucauldian Discourse Analysis, Thematic Analysis, and Qualitative Content Analysis. Second, we review the literature to show how these methods have been combined across disciplines and case studies. Third, we build upon existing literature to unpack the benefits and challenges of multi-method text/discourse analysis, and offer strategies to help navigate the problems that may arise. Overall, this article introduces multi-method qualitative text and discourse analysis (MMQTDA) as a methodological framework to provide guidance and offer solid foundations for an emerging methodological conversation in qualitative text research.