CoChat
Sign Up
Cited by 153Misinformation and Its ImpactsComplex Network Analysis TechniquesSocial Media and Politics

Spreading Disinformation on Facebook: Do Trust in Message Source, Risk Propensity, or Personality Affect the Organic Reach of “Fake News”?

Open access
Buchanan, Tom, Benson, Vladlena·2019·Social Media + Society
DOI 10.1177/2056305119888654
Open this paper in CoChat — free accountOpen PDF

Abstract

There is considerable concern about the propagation of disinformation through social media, particularly for political purposes. “Organic reach” has been found to be important in the propagation of disinformation on social networks. This is the phenomenon whereby social media users extend the audience for a piece of information: interacting with it, or sharing it with their wider networks, greatly increases the number of people the information reaches. This project evaluated the extent to which characteristics of the message source (how trustworthy they were) and the recipient (risk propensity and personality) influenced the organic reach of a potentially false message. In an online study, 357 Facebook users completed personality and risk propensity scales and rated their likelihood of interacting in various ways with a message posted by either a trustworthy or untrustworthy source. Message source impacted on overall organic reach, with messages from trusted sources being more likely to be propagated. Risk propensity did not influence reach. However, low scores on trait agreeableness predicted greater likelihood of interacting with a message. The findings provide preliminary evidence that both message source and recipient characteristics can potentially influence the spread of disinformation.

AI snapshot

Written from the paper’s full text.

TL;DR

In an online scenario study, Facebook users were more likely to say they would share, like, comment on, or react to a potentially fake news post when it came from a trusted close friend rather than an unknown contact. Risk propensity did not predict organic reach, while lower Agreeableness was associated with greater likelihood of propagating the post.

Key findings

  • The trust manipulation was effective: participants rated themselves more likely to trust information in the high-trust condition (M = 2.47, SD = 1.09) than in the low-trust condition (M = 1.69, SD = .92), t(349.65) = −7.34, p < .0005.
  • In multiple regression, trust condition significantly predicted organic reach (B = 1.55, SE = 0.36, β = .22, t = 4.31, p = .000), with higher reach for the higher-trust source.
  • Agreeableness had a significant negative effect on organic reach (B = −0.12, SE = 0.05, β = −.15, t = −2.60, p = .01): less agreeable people were more likely to extend the message’s reach.
  • Risk propensity was not significantly associated with organic reach (B = 0.00, SE = 0.02, β = .01, t = 0.20, p = .843), and the other four personality traits were non-significant; the model explained R² = .081, F = 4.33, p = .000.

Methods

  • Online study conducted via www.personalitytest.org.uk and Qualtrics; final sample was 409 participants, of whom 357 were Facebook users, with respondents from 44 countries and a majority relatively young female US students.
  • Measures included a 41-item Big Five personality questionnaire, Meertens and Lion’s 7-item Risk Propensity Scale, and an experimental trust manipulation: high-trust condition described a close friend’s Facebook post about political corruption, while low-trust condition described the same post from someone who had recently sent a friend request but was not really known.
  • Organic reach was measured as a summed scale of four self-rated likelihood items—share, like, comment, and react—with acceptable reliability (α = .80); analysis used a manipulation-check t-test and multiple regression predicting organic reach from trust condition, risk propensity, and the five personality variables.

Limitations

  • The paper states that only a single exemplar of a potential fake news item was used, so the Agreeableness effect may be specific to that item and requires replication with other materials; the election-campaign scenario also limits generalizability because trust is context dependent.
  • The paper states that the trust manipulation confounded trustworthiness with closeness of the relationship, that the truthfulness of the story was never made clear to participants, and that the study relied entirely on self-reported likelihood of interaction rather than actual behavioral observations.
  • The paper states that the measured variables accounted for only a small proportion of variance in self-rated likelihood of propagating disinformation (R² = .081), with unmeasured factors such as digital literacy, political interest, identity consistency, and general social media behavior likely playing a role; the convenience sample’s skew toward young female US students is also evident from the reported demographics.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1177/2056305119888654

Cites

28 of 28 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

0 of 25 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    A higher level of trust in the source of a Facebook message increases the likelihood that recipients will extend its organic reach (share, like, comment, react).

    “The analysis, summarized in Table 4, indicated there was a statistically significant effect of trust manipulation on organic reach, with a higher level of reach for posts in the higher trust condition.”
    • Also stated in: § Discussion · plain
      “As hypothesized, trustworthiness of the source of a message was associated with its potential organic reach.”
    • Also stated in: § Discussion · hedged
      “People rated themselves as more likely to extend the reach of a message coming from a trustworthy friend by sharing, liking, and so on.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: All four Facebook actions (share, like, comment, react) contribute to the organic reach of a posting.
  • Findingown data

    Participants in the high-trust condition reported a higher likelihood of trusting information posted by someone like that (M = 2.47, SD = 1.09) than participants in the low-trust condition (M = 1.69, SD = .92), a statistically significant difference.

    “As predicted, participants reported a higher likelihood of trusting the information in the high-trust condition (M = 2.47, SD = 1.09) than in the low-trust condition (M = 1.69,SD = .92).This difference was statistically significant (t (349.65)= -7.34,p < .0005,using adjusted degrees of freedom due to a significant Levene's test for equality of variance across conditions).”
    Quote in the paperSubstantial claim
  • Findingown data

    Among Facebook users, Agreeableness is negatively associated with organic reach: less agreeable people are more likely to increase a message's reach.

    “Beyond this, only Agreeableness had a statistically significant effect on organic reach.Agreeableness was negatively associated with reach: less agreeable people were more likely to increase the message's reach.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: All four Facebook actions (share, like, comment, react) contribute to the organic reach of a posting.
  • Findingown data

    Four of the five personality dimensions (Extraversion, Conscientiousness, Neuroticism, Openness) had no effect on potential organic reach.

    “Of the five dimensions included, four had no effect on the measure of potential organic reach”
    Quote in the paperSubstantial claim
  • Findingown data

    Risk propensity had no effect on the organic reach of fake news.

    “The hypothesized effect of risk propensity on organic reach of fake news was not found.”
    • Also stated in: § Conclusion · plain
      “Risk propensity | 0.00 | 0.02 | .01 | 0.20 | .843”
    Quote in the paperSubstantial claim
  • May combine two claims. 14
  • Classified as probably not a substantial claim. 22

  • May combine two claims.

    Findingown data

    Risk propensity and the other four personality traits (Extraversion, Neuroticism, Openness, Conscientiousness) did not have statistically significant effects on organic reach.

    “Beyond this, only Agreeableness had a statistically significant effect on organic reach.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: All four Facebook actions (share, like, comment, react) contribute to the organic reach of a posting.

  • May combine two claims.

    Findingcitation

    The effect size for trust on organic reach was β = .22, which exceeded the .2 threshold proposed for an effect that would have real-world significance.

    “At β = .22,the effect size exceeded the .2threshold proposed for an effect that would have realworld significance (Ferguson, 2009).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Ferguson, 2009

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    More agreeable people gave lower organic reach scores, indicating they were less likely to contribute to the propagation of a fake news item.

    “More agreeable people gave lower organic reach scores, and so were less likely to contribute to the propagation of a fake news item.”
    • Also stated in: § Conclusion · plain
      “Agreeableness | -0.12 | 0.05 | -.15 | -2.60* | .01”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductcitation

    The 41-item personality questionnaire has been validated for use on the internet and correlates well with the dimensions of Costa and McCrae's model.

    “This has been validated for use on the internet (Buchanan et al., 2005) and correlates well with the dimensions of Costa and McCrae's (1992) model.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Buchanan et al., 2005

      Not checked in the free run.

    • cites Costa and McCrae's (1992)

      Not checked in the free run.

  • May combine two claims.

    Study design and conductown data

    The trust manipulation was effective, and the two conditions differed in the extent to which the information source described was considered trustworthy by participants.

    “This indicates that the trust manipulation was effective, and the two conditions differed in the extent to which the information source described was considered trustworthy by participants.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Individuals who used social media had higher levels of risk acceptance than those who did not.

    “Fogel and Nehmad (2009) showed that individuals who used social media had higher levels of risk acceptance than those who did not.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Fogel and Nehmad (2009)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Qualtrics' proprietary technology was used to prevent multiple completions, and the dataset was screened for unrealistic combinations of demographic data.

    “Qualtrics' proprietary technology was used to prevent multiple completions, and the dataset was screened for unrealistic combinations of demographic data.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Social network users' perceptions of privacy risk influenced their attitude toward sharing information online; people who judged information sharing as being more risky had less positive attitudes toward it.

    “Hajli and Lin (2016) found that social network users' perceptions of privacy risk influenced their attitude toward sharing information online.People who judged information sharing as being more risky had less positive attitudes toward it.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hajli and Lin (2016)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Social media users with high levels of risk propensity are more likely to become victims of cybercrime.

    “There are indications that social media users with high levels of risk propensity are more likely to become victims of cybercrime (Saridakis et al., 2016;Whittle et al., 2013).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Saridakis et al., 2016

      Not checked in the free run.

    • cites Whittle et al., 2013

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    The finding that trust in the source is associated with organic reach is consistent with Williams et al.'s proposal that trust in a message source increases vulnerability to online influence.

    “This is consistent with Williams et al.'s (2017) proposal that trust in a message source increases our vulnerability to online influence, for example, in the context of internet scams.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Williams et al.'s (2017)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    An individual's status on all five personality dimensions can be inferred from their social media footprint.

    “There is evidence that an individual's status on all of these dimensions can be inferred from their social media footprint (e.g., Azucar et al., 2018;Hinds & Joinson, 2019).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Azucar et al., 2018

      Not checked in the free run.

    • cites Hinds & Joinson, 2019

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    In the context of sharing news information online, individuals are more likely to trust, and engage with, a story when it is shared by someone they have a higher level of trust in.

    “In the context of sharing news information online, there is evidence that individuals are more likely to trust, and engage with, a story when it is shared by someone they have a higher level of trust in (Sterret et al., 2018).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Sterret et al., 2018

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    The finding that trust in the source is associated with organic reach chimes with Sterret et al.'s finding that trust in the source of information influences the extent to which we interact with it.

    “It also chimes with Sterret et al.'s (2018) finding that trust in the source of information influences the extent to which we interact with it.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Sterret et al.'s (2018)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Extant research indicates that trust and risk propensity are among the factors that impact cyber-victimization of individuals.

    “Extant research indicates that trust and risk propensity are among the factors that impact cyber-victimization of individuals (Saridakis et al., 2016;Williams et al., 2017).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Saridakis et al., 2016

      Not checked in the free run.

    • cites Williams et al., 2017

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    Meertens and Lion's Risk Propensity Scale has acceptable reliability and has been used successfully in an online format.

    “It has acceptable reliability, and has been used successfully in an online format (Branley & Covey, 2017).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Branley & Covey, 2017

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductown data

    Participants who used Facebook were randomly assigned to see either the high-trust or the low-trust version of the scenario.

    “Of those who did, half were then randomly assigned to see the high-trust version of the scenario, and the other half saw the low-trust version.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    Generalizability is limited by the nature of the trust manipulation and the election campaign scenario used.

    “Generalizability is also limited by the nature of the trust manipulation and the scenario used (an election campaign).”
    • Also stated in: § Limitations · plain
      “The single scenario used (an election campaign) also limits generalizability.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Trust is context dependent, and who one trusts may differ across different scenarios.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Extraversion is related to the number of Facebook friends a user has, Agreeableness to the type of affect expressed in status updates, Conscientiousness to the presence of political material in status updates, Neuroticism to frequency of posting, and Openness to Experience to the sharing of various forms of media.

    “For example, Extraversion is related to the number of Facebook friends a user has, Agreeableness to the type of affect (positive or negative) expressed in status updates, Conscientiousness to the presence of political material in status updates, Neuroticism to frequency of posting, and Openness to Experience is related to the sharing of various forms of media (Hall et al., 2013).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hall et al., 2013

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Agreeableness is known to be detectible from social media footprints.

    “There is evidence that this would be feasible, as Agreeableness is known to be detectible from social media footprints (Azucar et al., 2018).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Azucar et al., 2018

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    People vary in the extent to which they are willing to accept risk in everyday life.

    “People are known to vary in the extent to which they are willing to accept risk in everyday life (Meertens & Lion, 2008).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Meertens & Lion, 2008

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Personality-targeted advertising on social media can influence user behavior.

    “As an example, work by Matz et al. (2017) has shown that personality-targeted advertising on social media can indeed influence user behavior.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Matz et al. (2017)

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Openness to Experience is positively associated, and Neuroticism negatively associated, with users' self-reports of the breadth of their self-disclosure on Facebook; Extraversion is positively associated with depth of self-disclosure.

    “In terms of informationsharing behavior, Openness to Experience is positively associated, and Neuroticism negatively associated, with users' self-reports of the breadth of their self-disclosure on Facebook.In addition, Extraversion is positively associated with depth of self-disclosure (Hollenbaugh & Ferris, 2014).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hollenbaugh & Ferris, 2014

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning

    The effect of Agreeableness on organic reach may be specific to the particular fake news item used and may not generalize to other types of material.

    “The current dataset does not allow us to say whether this effect would apply to all fake news items, or is particular to the item described here.”
    • Also stated in: § Limitations · hedged
      “It remains to be seen whether Agreeableness would also influence the organic reach of other types of material.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Trust in virtual communities and teams increases knowledge sharing and information flow.

    “It has been shown that trust in virtual communities and teams increases knowledge sharing and information flow (Watson-Manheim & Bélanger, 2007).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Watson-Manheim & Bélanger, 2007

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    In 2018, Twitter identified fraudulent accounts set up by a group linked to disinformation, which simulated those of US local newspapers.

    “It is pertinent to note that in 2018, Twitter identified fraudulent accounts set up by a group linked to disinformation, which simulated those of US local newspapers (Mak & Berry, 2018).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mak & Berry, 2018

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    The effect size of Agreeableness is low, but at massive social media scale it has potential real-world significance.

    “While the effect size is low, when operating at the massive scale made possible by social media, the effect has potential real-world significance.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    An estimated 10% of Facebook users spread such information.

    “The beliefs and motivations of the estimated 10% of Facebook users who do spread such information (Guess et al., 2019) require examination.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Guess et al., 2019

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Believing a story was false was not necessarily a barrier to sharing it.

    “It is worth noting a Pew Research Center survey that found believing a story was false was not necessarily a barrier to sharing it (Barthel et al., 2016).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Barthel et al., 2016

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    The present findings point to potential strategies for spreading disinformation.

    “The present findings point to potential strategies for spreading disinformation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    The fraudulent accounts appear to have been sleeper accounts established specifically for the purpose of building trust.

    “These appear to have been sleeper accounts established specifically for the purpose of building trust, just as envisioned in this scenario.”
    Quote in the paperNot a substantial claim
30 of 30

  • The authors state this as an assumption.

    Users who encounter disinformation online and spread it further are victims, duped by the message originators into spreading lies.

    “Those who encounter the material online and spread it further are essentially victims, duped by the message originators into spreading lies.”

  • The paper relies on this without support.

    Disinformation is, by its very nature, designed to provoke controversy.

    “Disinformation is, by its very nature, designed to provoke controversy.”

  • The paper relies on this without support.

    All four Facebook actions (share, like, comment, react) contribute to the organic reach of a posting.

    “While Facebook's algorithms assign different weights to each of these user behaviors in determining whether a user's audience sees the post they are responding to, it is clear that all contribute to the organic reach of a posting.”

  • The paper relies on this without support.

    Trust is context dependent, and who one trusts may differ across different scenarios.

    “Trust is very much context dependent, and who one trusts may differ across different scenarios.”

  • The paper relies on this without support.

    In real life, individuals often have no way of knowing for sure whether information they come across is true.

    “In fact, this would normally be the situation in real life: individuals often have no way of knowing for sure whether information they come across is true.”
5 of 5

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?

  • Trust in information on social media and trust in the platforms themselves positively influence people's intention to share information, news, and opinions on social media.

    “All the studies in our corpus suggest that trust in the information on social media [10,56,65,68] and trust in the platforms themselves [30,59,66,70] positively infuence people's intention to share information, news, and opinions on social media [122].”

    From What Do We Mean When We Talk about Trust in Social Media? A Systematic Review (2023) · cites this paper as [10]

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Consistentt(349.65) = -7.34reported p < .0005 · recomputed p < .001

    “t (349.65)= -7.34,p < .0005”

1 of 1

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Self-check results appear here once an independent check has confirmed them.

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. Costa P. T., 1992, Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO FFI): Professional manual

Cited by

Sorted by citations

Related

Authors
Buchanan, T., & Benson, V.
Affiliations
  • University of Westminster (GB)
  • Aston University (GB)
Corresponding author
Buchanan, Tom
ORCID
Venue
Social Media + Society
Publisher
SAGE Publications
Published
2019-10
Volume
5
Issue
4
ISSN
2056-3051 (ISSN-L)
DOI
10.1177/2056305119888654
OpenAlex
W2983472018
Semantic Scholar
12c371b58a7a5083d9451ceac712078d92dd927a
Cited by
151
References
28
Citations by year
2026: 13 · 2025: 17 · 2024: 22 · 2023: 40 · 2022: 27 · 2021: 19 · 2020: 12 · 2019: 1
FWCI
29.63
Type
journal article
Language
en
Open access
Licence
Locations
  • Social Media + Society (publishedVersion, cc-by-nc)
  • WestminsterResearch (University of Westminster) (acceptedVersion, cc-by-nc)
  • Aston Publications Explorer (Aston University) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
  • Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Complex Network Analysis Techniques
  • Social Sciences › Communication › Social Media and Politics
Keywords
disinformation, fake news, organic reach, Facebook, risk propensity, personality, agreeableness, information sharing, political disinformation

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.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-10.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-10.

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