Last Date for Paper Submission: 30th April, 2026

Filter Bubbles and Epistemic Isolation: Psychological Mechanisms and Measurement in Algorithmically Curated News

Author Name: Mrs. Sonal Agarwal Date: 28-03-2026

Filter bubbles personalized information environments in which algorithmic curation systematically narrows the diversity of perspectives, sources, and topics to which users are exposed represent a theoretically distinctive concept from echo chambers, yet the two are routinely conflated in both academic literature and public discourse. This paper provides the first systematic conceptual analysis distinguishing filter bubbles (a structural property of information environments) from echo chambers (a psychological property of belief systems), and develops a measurement framework for each. Drawing on information diversity theory, selective exposure research, and Pariser’s (2011) original formulation, the paper proposes the Information Environment Diversity Index (IEDI) a multidimensional metric assessing source diversity, viewpoint diversity, topic diversity, and temporal diversity in algorithmically curated news feeds — as the appropriate measure of filter bubble intensity. The paper reviews empirical evidence on the prevalence of filter bubbles from audit studies, computational analysis of platform recommendation patterns, and user experience surveys, synthesizing the counterintuitive finding that measured filter bubbles are often weaker than theorized and that individual choice behaviors contribute more to perspective limitation than algorithmic curation. The psychological mechanisms connecting filter bubble exposure to cognitive and attitudinal outcomes are evaluated: selective exposure reinforcement, availability heuristic distortion, perceived consensus inflation, and epistemic overconfidence. A randomized experiment design for testing the causal effects of filter bubble intensity on psychological outcomes is proposed, incorporating the IEDI measurement framework alongside validated psychological outcome measures.

Keywords: filter bubbles; information diversity; selective exposure; epistemic overconfidence; algorithmic curation; personalization; news diversity; media psychology.

Description of the logo Download PDF

Scroll to Top