Automating the News: How Personalized News Recommender System Design Choices Impact News Reception
Abstract
This study investigates the impact of personalized news recommender system design on selective exposure, elaboration, and knowledge. Scholars have worried that proliferation of personalization technologies will degrade public opinion by isolating people from challenging perspectives. Informed by selective exposure research, this study examines personalized news recommender system designs using a communication mediation model. Recommender system design choices examined include computer-generated personalized recommendations, user-customized recommendations, and full or limited news information environments based on recommendations. Results from an online mock election experiment with Ohio adult Internet users indicate increased selective exposure when using personalized news systems. However, portals recommending news based on explicit user customization result in significantly higher counterattitudinal news exposure. Expected positive effects on elaboration and indirect effects on knowledge through elaboration are found only in personalized news recommender systems that display only recommended headlines. Lastly, personalized news recommender system use has a negative direct effect on knowledge.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Existing evidence suggests that the pattern of individuals' information seeking and consumption is more nuanced and multifaceted than assumed in the notion of selective exposure and homophily.
“Existing evidence suggests that the pattern of individuals' information seeking and consumption is more nuanced and multifaceted than assumed in the notion of selective exposure and homophily.”
From Do Search Algorithms Endanger Democracy? An Experimental Investigation of Algorithm Effects on Political Polarization (2020) · cites this paper as Beam, 2014
People generally do not necessarily avoid counter-attitudinal information, and further seek such information when they believe it helps them understand and be accurate on the situation at hand.
“People, in general, do not necessarily avoid counter-attitudinal information (Beam, 2014;Garrett et al., 2013), and further seek such information when they believe it helps them understand and be accurate on the situation at hand (Hart et al., 2009;Holbert et al., 2013;Kunda, 1990).”
From Do Search Algorithms Endanger Democracy? An Experimental Investigation of Algorithm Effects on Political Polarization (2020) · cites this paper as Beam, 2014
No citation lists for this paper
No citation lists for this paper yet — checked again daily.
References 217 · Cited by 159
- Authors
- Beam, M. A.
- Affiliations
- Washington State University (US)
- Corresponding author
- Beam, Michael A.
- ORCID
- Beam, Michael A.
- Venue
- Communication Research
- Publisher
- SAGE Publications
- Published
- 2013-07-31
- Volume
- 41
- Issue
- 8
- Pages
- 1019-1041
- ISSN
- 0093-6502 (ISSN-L)
- 1552-3810
- OpenAlex
- W2134021902
- Cited by
- 159
- References
- 217
- Citations by year
- 2026: 12 · 2025: 16 · 2024: 9 · 2023: 16 · 2022: 14 · 2021: 15 · 2020: 21 · 2019: 15 · 2018: 15 · 2017: 13 · 2016: 5 · 2015: 7 · 2014: 1
- FWCI
- 13.42
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Communication Research (publishedVersion)
- Topics
- Social Sciences › Sociology and Political Science › Media Influence and Politics
- Physical Sciences › Computer Science › Information Systems › Recommender Systems and Techniques
- Social Sciences › Communication › Media Studies and Communication
- Keywords
- personalized news, news recommender systems, selective exposure, elaboration, public opinion, echo chambers, personalization, news websites
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
