Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 83 of the 86 references it cites were found in a registry or an index.

Opinion Dynamics with Confirmation Bias

Open access

Abstract

BACKGROUND: Confirmation bias is the tendency to acquire or evaluate new information in a way that is consistent with one's preexisting beliefs. It is omnipresent in psychology, economics, and even scientific practices. Prior theoretical research of this phenomenon has mainly focused on its economic implications possibly missing its potential connections with broader notions of cognitive science. METHODOLOGY/PRINCIPAL FINDINGS: We formulate a (non-Bayesian) model for revising subjective probabilistic opinion of a confirmationally-biased agent in the light of a persuasive opinion. The revision rule ensures that the agent does not react to persuasion that is either far from his current opinion or coincides with it. We demonstrate that the model accounts for the basic phenomenology of the social judgment theory, and allows to study various phenomena such as cognitive dissonance and boomerang effect. The model also displays the order of presentation effect-when consecutively exposed to two opinions, the preference is given to the last opinion (recency) or the first opinion (primacy) -and relates recency to confirmation bias. Finally, we study the model in the case of repeated persuasion and analyze its convergence properties. CONCLUSIONS: The standard Bayesian approach to probabilistic opinion revision is inadequate for describing the observed phenomenology of persuasion process. The simple non-Bayesian model proposed here does agree with this phenomenology and is capable of reproducing a spectrum of effects observed in psychology: primacy-recency phenomenon, boomerang effect and cognitive dissonance. We point out several limitations of the model that should motivate its future development.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0099557

Cites

83 of 83 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Opinions described via Gaussian densities (17). The initial opinion of is described by Gaussian probability density p(x) (blue curve) centered at zero; see (17). The opinion of amounts to Gaussian probability density q(x) (purple curve) centered at a positive value. For all three figures continuous density f(x) () were approximated by 100 points , . The resulting opinion of is given by (16) with (olive curve). (a) The opinion of moves towards that of ; , , , . (b) The maximally probable opinion of is reinforced; , , , . (c) The change of the opinion of is relatively small provided that the Gaussian densities overlap only in the region of non-commitment; cf. (18), (19). Whenever the densities overlap only within the rejection range the difference between p(x) and is not visible by eyes. For example, if p(x) and q(x) are Gaussian with, respectively, , , , the Hellinger distance (see (30) for definition) is close to maximally far, while the opinion change is small: .
Figure 1 Opinions described via Gaussian densities (17). The initial opinion of is described by Gaussian probability density p(x) (blue curve) centered at zero; see (17). The opinion of amounts to Gaussian probability density q(x) (purple curve) centered at a positive value. For all three figures continuous density f(x) () were approximated by 100 points , . The resulting opinion of is given by (16) with (olive curve). (a) The opinion of moves towards that of ; , , , . (b) The maximally probable opinion of is reinforced; , , , . (c) The change of the opinion of is relatively small provided that the Gaussian densities overlap only in the region of non-commitment; cf. (18), (19). Whenever the densities overlap only within the rejection range the difference between p(x) and is not visible by eyes. For example, if p(x) and q(x) are Gaussian with, respectively, , , , the Hellinger distance (see (30) for definition) is close to maximally far, while the opinion change is small: .CC BY · PubMed Central
Figure 2 Opinions described via bump densities (28). Blue curve: the initial opinion of given by (28) with b = 1. Purple curve: the opinion of described by (28) with . Olive curve: the resulting opinion of obtained via (16) with .
Figure 2 Opinions described via bump densities (28). Blue curve: the initial opinion of given by (28) with b = 1. Purple curve: the opinion of described by (28) with . Olive curve: the resulting opinion of obtained via (16) with .CC BY · PubMed Central
Figure 3 Opinion change versus discrepancy. (a) The opinion change is quantified via the Hellinger distance between the old and new opinion of (blue curves); see (30) for the definition. For comparison we also include the total variance distance (purple curves); see (33). These two distances are plotted versus the discrepancy . The initial opinion of the agent is Gaussian with and ; see (17). The opinion of is Gaussian with and . Thus m quantifies the initial distance between the opinions of and . The final opinion is given by (13). Different curves correspond to different . Blue curves: for (upper curve) and (lower curve). Purple curves: for (upper curve) and (lower curve). The maximum of h(m) () is reached at (). (b) () is the point where h(m) () achieves its maximum as a function of m. Blues points: versus for same parameters as in (a). grows both for and , e.g. , , , . Purple points: versus for same parameters as in (a). (c) The difference of the anchors (maximally probable values) versus for the initial opinions of and given by (17) under , , and . The final opinion of (and its maximally probable value ) if found from (13) under (black points), (blue points) and (red points).
Figure 3 Opinion change versus discrepancy. (a) The opinion change is quantified via the Hellinger distance between the old and new opinion of (blue curves); see (30) for the definition. For comparison we also include the total variance distance (purple curves); see (33). These two distances are plotted versus the discrepancy . The initial opinion of the agent is Gaussian with and ; see (17). The opinion of is Gaussian with and . Thus m quantifies the initial distance between the opinions of and . The final opinion is given by (13). Different curves correspond to different . Blue curves: for (upper curve) and (lower curve). Purple curves: for (upper curve) and (lower curve). The maximum of h(m) () is reached at (). (b) () is the point where h(m) () achieves its maximum as a function of m. Blues points: versus for same parameters as in (a). grows both for and , e.g. , , , . Purple points: versus for same parameters as in (a). (c) The difference of the anchors (maximally probable values) versus for the initial opinions of and given by (17) under , , and . The final opinion of (and its maximally probable value ) if found from (13) under (black points), (blue points) and (red points).CC BY · PubMed Central
Figure 4 Order of presentation effect. Blue curve: The initial opinion of is described by Gaussian probability density p(x) with and ; see (17). Purple (resp. olive) curve: the initial opinion of (resp. ) are given by (17) with (resp. ) and (resp. ). Green curve: the resulting opinion of after interacting first with and then with . Both interactions use . The final opinion of is inclined to the most recent opinion (that of ) both with respect to its maximally probable value and distance. The final opinion of has a larger width than the initial one.
Figure 4 Order of presentation effect. Blue curve: The initial opinion of is described by Gaussian probability density p(x) with and ; see (17). Purple (resp. olive) curve: the initial opinion of (resp. ) are given by (17) with (resp. ) and (resp. ). Green curve: the resulting opinion of after interacting first with and then with . Both interactions use . The final opinion of is inclined to the most recent opinion (that of ) both with respect to its maximally probable value and distance. The final opinion of has a larger width than the initial one.CC BY · PubMed Central
Figure 5 Cognitive dissonance. (a) Blue (resp. purple) curve: the initial opinion of agent (resp. ) described by probability density p(x) (resp. q(x)). Olive curve: the final opinion of as given by (16) with . Here p(x) and q(x) are defined by (17) with , , , . The final opinion develops two peaks of comparable height (cognitive dissonance). (b) Avoiding the cognitive dissonance due to a larger : the second peak is much smaller (other parameters are those of (a)). (c) Avoiding the cognitive dissonance due to a smaller : the first peak is much smaller (other parameters are those of (a)).
Figure 5 Cognitive dissonance. (a) Blue (resp. purple) curve: the initial opinion of agent (resp. ) described by probability density p(x) (resp. q(x)). Olive curve: the final opinion of as given by (16) with . Here p(x) and q(x) are defined by (17) with , , , . The final opinion develops two peaks of comparable height (cognitive dissonance). (b) Avoiding the cognitive dissonance due to a larger : the second peak is much smaller (other parameters are those of (a)). (c) Avoiding the cognitive dissonance due to a smaller : the first peak is much smaller (other parameters are those of (a)).CC BY · PubMed Central
Figure 6 Opinion change in the boomerang regime. Blue (resp. purple) curve: the initial opinion of agent (resp. ) described by probability density p(x) (resp. q(x)). Olive curve: the final opinion of given by (16) with . Here p(x) and q(x) are given by (17) with and . The anchor (maximally probable opinion) of not only moves away from the anchor of ; but it is also enhanced: the (biggest) peak of is larger than that of p(x). The second (smaller) peak of arises because the initial probability of located to the right from the anchor of , moves away from ; gets a local minimum close to .
Figure 6 Opinion change in the boomerang regime. Blue (resp. purple) curve: the initial opinion of agent (resp. ) described by probability density p(x) (resp. q(x)). Olive curve: the final opinion of given by (16) with . Here p(x) and q(x) are given by (17) with and . The anchor (maximally probable opinion) of not only moves away from the anchor of ; but it is also enhanced: the (biggest) peak of is larger than that of p(x). The second (smaller) peak of arises because the initial probability of located to the right from the anchor of , moves away from ; gets a local minimum close to .CC BY · PubMed Central
Figure 7 Order of presentation effect in the boomerang regime. The same as in Fig. 4 but for (boomerang regime). Now the final opinion of is inclined to the first opinion (that of ) with respect to the distance. The initial maximally probable opinion of is still maximally probable. Moreover, its probability has increased and the width around it has decreased. The final opinion has 3 peaks.
Figure 7 Order of presentation effect in the boomerang regime. The same as in Fig. 4 but for (boomerang regime). Now the final opinion of is inclined to the first opinion (that of ) with respect to the distance. The initial maximally probable opinion of is still maximally probable. Moreover, its probability has increased and the width around it has decreased. The final opinion has 3 peaks.CC BY · PubMed Central
Figure 8 Illustration of the order of presentation effect in the boomerang regime. given by (50, 51) versus .
Figure 8 Illustration of the order of presentation effect in the boomerang regime. given by (50, 51) versus .CC BY · PubMed Central
Figure 9 Repeated persuasion in the boomerang regime. Blue (resp. purple) curve: the initial opinion of agent (resp. ) described by probability density p(x) (resp. q(x)) as given by (17) with , , . Olive curve: the opinion of after 50 iterations (52) with .
Figure 9 Repeated persuasion in the boomerang regime. Blue (resp. purple) curve: the initial opinion of agent (resp. ) described by probability density p(x) (resp. q(x)) as given by (17) with , , . Olive curve: the opinion of after 50 iterations (52) with .CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

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

The self-check needs the paper’s full text.

Citations

Cites

Show all 86 references

Research works printed but not linked: 3

  1. CE Alchourrón, 1985, On the logic of theory change. J. Symb, Logic, 50, 510
  2. P beim Graben, 2006, Pragmatic information in dynamic semantics, Mind and Matter, 4, 169
  3. GJ Szekely, 2005, Half of a coin: negative probabilities, Wilmott Magazine, 50, 66

Cited by

See all 90

Metadata

Authors
Allahverdyan, A. E., & Galstyan, A.
Affiliations
  • A. Alikhanyan National Laboratory (AM)
  • University of Southern California (US)
Corresponding author
Allahverdyan, Armen E.
Venue
PLoS ONE
Published
2014-07-09
Volume
9
Issue
7
Pages
e99557
ISSN
1932-6203 (ISSN-L)
OpenAlex
W2045209926
Cited by
90
References
86
Citations by year
2026: 3 · 2025: 8 · 2024: 12 · 2023: 16 · 2022: 8 · 2021: 10 · 2020: 9 · 2019: 5 · 2018: 4 · 2017: 6 · 2016: 7 · 2015: 2
FWCI
3.29
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • arXiv (Cornell University)
Funders
  • Defense Advanced Research Projects Agency (W911NF- 12-1-0034, No. W911NF-12-1-0034)
  • Multidisciplinary University Research Initiative (FA9550-10-1-0569)
  • Air Force Office of Scientific Research (FA9550-, FA9550-10-1-, FA9550-10-1-0569)
Topics
  • Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Opinion Dynamics and Social Influence
  • Social Sciences › Decision Sciences › General Decision Sciences › Decision-Making and Behavioral Economics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Cognitive Science and Education Research
Keywords
confirmation bias, opinion dynamics, persuasion, subjective probability, social judgment theory, cognitive dissonance, boomerang effect, primacy effect, recency effect, presentation order
MeSH terms
Algorithms, Cognition, Humans, Models, Psychological

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Opinion Dynamics with Confirmation Bias. https://app.cochat.ai/papers/10.1371/journal.pone.0099557
The paper
Allahverdyan, A. E., & Galstyan, A. (2014). Opinion Dynamics with Confirmation Bias. PLoS ONE. https://doi.org/10.1371/journal.pone.0099557

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.