The DOI matches its Crossref record; 125 of the 125 references it cites were found in a registry or an index; 43 of the 130 sources its claims cite were checked for support.
Social Drivers and Algorithmic Mechanisms on Digital Media.
Open accessPublished version of a preprint
Abstract
Algorithms on digital media are powered by social drivers, creating a feedback loop; existing evidence suggests algorithms mostly reinforce existing social drivers.
Overview
CoChat Summary
## TL;DR
The paper argues that digital-media algorithms should not be treated as standalone causes of harm: they operate in feedback loops with basic social motives, platform design, and broader societal context. Across well-being, misinformation, and polarization, the evidence it reviews suggests algorithms usually reinforce existing social drivers rather than independently create the problems.
## Key findings
- The paper states that direct evidence that algorithms harm well-being is “scarce,” and that existing studies often cannot separate algorithm effects from general platform use, social behavior, or societal change.
- For mental health, it highlights one algorithm-specific result: reducing positive posts in Facebook’s feed reduced the likelihood of users posting positive content by 0.1% (Kramer et al., 2014).
- For well-being more broadly, it reports that Facebook deactivation trials found small to moderate decreases in anxiety, one decrease in depression, one decrease in loneliness, and that life satisfaction did not change after 1 week but increased after 4 weeks in one study.
- For polarization and misinformation, it says online echo chambers are likely smaller than offline echo chambers, misinformation is a small proportion of digital-news consumption and is shared by a tiny minority of users, and a YouTube study found only 1 out of 100,000 users who started with moderate content later moved to far-right content.
## Methods
- Narrative brief overview/review of the current evidence on algorithms’ roles in well-being, misinformation, and polarization.
- Synthesizes findings from prior studies, including randomized controlled trials, natural experiments, field experiments, experience-sampling studies, and platform-data analyses.
- Uses the reviewed evidence to build a conceptual argument about social drivers, algorithmic mechanisms, and broader societal context.
## Limitations
- The paper itself notes that most existing studies cannot distinguish algorithm effects from general social-media use, social behavior, or societal change because these traces are intermingled in the data.
- It explicitly says evidence remains scarce and that substantial further empirical research is needed.
- Evident from the text: it is a brief overview, not a new empirical study, so its claims depend on the quality and scope of prior work it reviews.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
31 Valid · 3 Not valid · 42 Not checked · 27 Own results and interpretations
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“Most movement to far-right videos comes from outside the platform”
§ Social Drivers and Algorithmic Mechanisms on Digital Media > Social Drivers and Algorithmic Mechanisms Influencing the Collective Dynamics of Political Polarization and Misinformation
What the sources say
- Disputedcites Hosseinmardi et al., 2021 sourceExperimental — accuracy not yet measured
The Results (p. 7) report that 41% of referrals to far-right videos came from external URLs, not a majority. The larger figure of nearly 55% also includes the YouTube homepage and search queries, some of which were made inside YouTube; the Discussion (p. 8) describes these collectively as sources other than a previous YouTube video, not as sources outside the platform. (LLM, conf=0.88)
Checked by a reasoning model — may be wrong.
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Disputed
Some cited sources support this claim and some don’t — read each source’s verdict below.
In the paper
“Additionally, misinformation has been shown not to easily change beliefs or political voting behavior (Bail et al., 2020; Guess et al., 2020).”
§ Social Drivers and Algorithmic Mechanisms on Digital Media > Social Drivers and Algorithmic Mechanisms Influencing the Collective Dynamics of Political Polarization and Misinformation
What the sources say
- Partly supportedcites Bail et al., 2020 sourceExperimental — accuracy not yet measured
The Conclusion reports no substantial effect of interactions with Russian IRA accounts on political attitudes among frequent partisan Twitter users in late 2017. But it explicitly says the study could not assess voting behavior, and its findings do not establish the broader claim for misinformation generally. (LLM, conf=0.90)
- Disputedcites Guess et al., 2020 sourceExperimental — accuracy not yet measured
Finding 5 reports that randomized exposure to a dubious article increased belief in its central false claim. Finding 6 reports increased intent to vote, contrary to the cited sentence’s claim about resistance to change, although the paper measured voting intention rather than actual voting behavior. (LLM, conf=0.87)
Checked by a reasoning model — may be wrong.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Social media and its algorithms are so successful because they build on ancient human needs for connection and status”
- Also stated in: § Social Drivers Underlying Individual and Group Behavior on Digital Media · plain
“Social media and its algorithms are so successful because they build on ancient human needs for connection and status (Brady et al., 2020; Meshi et al., 2015; Nadkarni & Hofmann, 2012).”
§ Social Drivers and Algorithmic Mechanisms on Digital Media > Social Drivers Underlying Individual and Group Behavior on Digital Media
What the sources say
- Not checked — no full text of the sourcecites Brady et al., 2020 source
- Not checked — no full text of the sourcecites Meshi et al., 2015 source
- Partly supportedcites Nadkarni & Hofmann, 2012 sourceExperimental — accuracy not yet measured
The Concluding Comments (page unavailable) say Facebook use is driven by the needs to belong and to present oneself; the Need to Belong section also discusses social standing. This supports the connection and status aspects, but the paper does not establish that algorithms exploit these needs or that doing so explains their success. It does not characterize the needs as ancient. (LLM, conf=0.87)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § Social Drivers Underlying Individual and Group Behavior on Digital Media · plain
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Most current digital-media algorithms strongly optimize for engagement (Narayanan, 2023b; Nikolov et al., 2019).”
§ Social Drivers and Algorithmic Mechanisms on Digital Media > Algorithmic Mechanisms and Other Platform Influences
What the sources say
- Not matchedcites Narayanan, 2023b
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Partly supportedcites Nikolov et al., 2019 sourceExperimental — accuracy not yet measured
The Introduction (p. 3) says Web-platform algorithms steer attention toward engaging content and that some filtering depends on engagement metrics. It does not establish that most current digital-media algorithms strongly optimize for engagement; it also names relevance, timeliness, reputation, and user preferences as factors. (LLM, conf=0.86)
states a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Not matched
Every source the paper cites for this claim was checked and supports it.
In the paper
“For instance, users are more likely to like and share low-quality content that others have already liked (Avram et al., 2020).”
§ Social Drivers and Algorithmic Mechanisms on Digital Media > Algorithmic Mechanisms and Other Platform Influences
What the sources say
- Supportedcites Avram et al., 2020 source
The Implications section reports that, in the simulated Fakey feed, higher displayed social engagement was associated with higher liking/sharing rates for articles from low-credibility sources. It also interprets the result as showing that people are more likely to like or share such content when they see that others have engaged with it. (LLM, conf=0.91)
hedgedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
49 more claims · 12 with their cited sources checked
Create a free account to see all 103 claims and to verify the 3 references still open.
- It may state two separate claims as one. 36
- Probably not a substantial claim. 24
Citations
Cites
null (2016) · Philosophy Study · doi:10.17265/2159-5313/2016.09.003 · Cited by 28,317
I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience
Marwick, A. E., & boyd, D. (2011) · New Media & Society · doi:10.1177/1461444810365313 · Cited by 4,612
Universal dimensions of social cognition: warmth and competence
Fiske, S. T., Cuddy, A. J. C., & Glick, P. (2007) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2006.11.005 · Cited by 4,567 · cited 2×
“Status and connection are pivotal to explaining social behavior (Abele & Wojciszke, 2013; Fiske et al., 2007; Gurtman, 2009) across many domains.”
BackgroundExperimental evidence of massive-scale emotional contagion through social networks
Kramer, A. D. I., Guillory, J. E., & Hancock, J. T. (2014) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1320040111 · Cited by 3,687 · cited 1×
“Only one study pinpointed algorithmic effects, finding that reducing positive posts in the Facebook feed reduces the likelihood of users posting positive content by 0.1% (Kramer et al., 2014).”
BackgroundInfluentialExposure to ideologically diverse news and opinion on Facebook
Bakshy, E., Messing, S., & Adamic, L. A. (2015) · Science · doi:10.1126/science.aaa1160 · Cited by 3,245 · cited 2×
“Regarding specific algorithm effects, a study on Face-book data in 2014 (Bakshy et al., 2015) found that users’ social networks determined posts in their feeds much more strongly than the algorithm.”
BackgroundMethodExperimental Study of Inequality and Unpredictability in an Artificial Cultural Market
Salganik, M. J., Dodds, P. S., & Watts, D. J. (2006) · Science · doi:10.1126/science.1121066 · Cited by 2,231 · cited 1×
“However, social success and quality of content are only partly correlated (Salganik et al., 2006).”
BackgroundThe Handbook of evolutionary psychology
(2006) · Choice Reviews Online · doi:10.5860/choice.44-0622 · Cited by 2,121
The influence of social hierarchy on primate health
Sapolsky, R. M. (2005) · Science · doi:10.1126/science.1106477 · Cited by 2,113 · cited 1×
“The twin desires to get along and to get ahead are basic human motives that were crucial for survival in our ancestral environment in social groups (Cummins, 2005; Sapolsky, 2005).”
BackgroundTrump, Brexit, and the Rise of Populism: Economic Have-Nots and Cultural Backlash
Inglehart, R., & Norris, P. (2016) · SSRN Electronic Journal · doi:10.2139/ssrn.2818659 · Cited by 1,915
Exposure to opposing views on social media can increase political polarization
Bail, C. A., Argyle, L. P., Brown, T. W., Bumpus, J. P., Chen, H., Hunzaker, M. B. F., Lee, J., Mann, M., Merhout, F., & Volfovsky, A. (2018) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1804840115 · Cited by 1,863 · cited 1×
“Supporting this, a field experiment on U.S. Twitter observed increased political polarization after exposure to posts from opinion leaders of the opposing party (Bail et al., 2018) and experience sampling reveals consistent results (de Mello et al., 2022).”
Background
Research works printed but not linked: 1
- Guess A. M., Nyhan B., Lyons B., Reifler J. (2018). Avoiding the echo chamber about echo chambers: Why selective exposure to like-minded political news is less prevalent than you think (No. 2, pp. 1–25). Knight Foundation.
Cited by
Algorithm-mediated social learning in online social networks
Brady, W. J., Jackson, J. C., Lindström, B., & Crockett, M. J. (2023) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2023.06.008 · Cited by 132
“Private companies may never develop algorithms that fully align with human cooperation and problem-solving, because these companies are motivated by pro fi t [122,123].”
Halkiopoulos, C., Gkintoni, E., Aroutzidis, A., & Antonopoulou, H. (2025) · Diagnostics · doi:10.3390/diagnostics15040456 · Cited by 120
Understanding Social Media Addiction: A Deep Dive
Amirthalingam, J., & Khera, A. (2024) · Cureus · doi:10.7759/cureus.72499 · Cited by 122
“Infinite scrolling makes it simple for users to lose track of time and engage in prolonged platform sessions by doing away with natural stopping points like page breaks [1,12].”
Social Media and Youth Mental Health: Scoping Review of Platform and Policy Recommendations
Chhabra, J., Pilkington, V., Benakovic, R., Wilson, M. J., Sala, L. L., & Seidler, Z. E. (2025) · Journal of Medical Internet Research · doi:10.2196/72061 · Cited by 62
“Rapid advancements in the development and adoption of content recommender systems (a platform design that organizes, prioritizes, and promotes content into users’ feeds based on user actions, preferences, and demographics [48]) have also raised concerns [49].”
Zhang, K., Li, Z., Lin, Y., Guo, Y., Huang, R., Liu, G., & Zhang, B. (2025) · Scientific Reports · doi:10.1038/s41598-025-21535-z · Cited by 41
“Platform recommendation algorithms may influence the visibility of different content types, potentially affecting how health information from various sources reaches users 49,50 .”
Metadata
- Authors
- Metzler, H., & Garcia, D.
- Affiliations
- Institute for Globally Distributed Open Research and Education (SE)
- Complexity Science Hub Vienna, Austria
- Corresponding author
- Metzler, Hannah
- Venue
- Perspectives on Psychological Science
- Publisher
- SAGE Publications
- Published
- 2023-07-19
- Volume
- 19
- Issue
- 5
- Pages
- 735-748
- ISSN
- 1745-6916 (ISSN-L)
- 1745-6924
- PMC
- PMC11373151
- OpenAlex
- W4384665338
- Semantic Scholar
- 2b23d685d6902bb94ab11b2292530ddf34fc957c
- Cited by
- 204
- References
- 117
- Citations by year
- 2026: 144 · 2025: 46 · 2024: 10 · 2023: 4
- FWCI
- 10.29
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Locations
- Perspectives on Psychological Science (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- KOPS (University of Konstanz) (submittedVersion, cc-by)
- PsyArXiv (OSF Preprints) (submittedVersion)
- PsyArXiv (OSF Preprints) (submittedVersion, cc-by)
- Funders
- Vienna Science and Technology Fund (ICT20-028, VRG16-005)
- Topics
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Communication › Social Media and Politics
- Keywords
- algorithmic curation, social media algorithms, well-being, misinformation, political polarization, social drivers, feedback loops, individualism, populist politics, climate change, algorithmic design, algorithmic media
- MeSH terms
- Algorithms, Communication, Humans, Social Media
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Social Drivers and Algorithmic Mechanisms on Digital Media. https://app.cochat.ai/papers/10.1177/17456916231185057
- The paper
- Metzler, H., & Garcia, D. (2024). Social Drivers and Algorithmic Mechanisms on Digital Media. Perspectives on Psychological Science. https://doi.org/10.1177/17456916231185057
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.