“Algorithmic Personalization of Source Cues in the Filter Bubble:
Self-Esteem and Self-Construal Impact Information Exposure”
by Silvia Knobloch-Westerwick & Axel Westerwick, to be published in NEW MEDIA & SOCIETY
How do users pick out online information sources? Building on a self-regulation perspective to media use, this study investigates routes to self-enhancement (i.e., state self-esteem increase, SSE) through selective exposure to sources of political online information. Personal-self and social-self importance were conceptualized as moderators of self-enhancement. An experiment mimicked the filter bubble, as participants (n = 88) browsed only attitude-aligned political content. The experiment varied source cues, with two (of eight) bylines displaying individual participants’ name initials as author initials. The selective exposure time participants spent on messages from same-initials authors was logged to capture egotism (based on the well-established name-letter effect). Pre-exposure state self-esteem influenced self-enhancement, contingent upon both personal-self and social-self importance. Perceived source similarity affected post-exposure state self-esteem, contingent upon the same moderators. The findings show that algorithms can personalize source cues to attract users and impact self-esteem.