Abstract

Despite the potential benefits brought by algorithm recommendation, it also causes some disadvantages such as information cocoon. As a countermeasure, social media users may manipulate algorithm to meet their needs. However, prior studies on algorithm manipulation are either qualitative or focus on net effects without considering joint effects. Thus, this study theorizes and empirically tests a configurational model of algorithm manipulation by considering the effects of personality (i.e., openness), information quality (e.g., information narrowing and information redundancy), and algorithm quality (e.g., algorithm fairness, accountability, and transparency). The fuzzy-set qualitative comparative analysis (fsQCA) based on survey data confirms our hypotheses.

Download

Share

COinS