Figure 2 - uploaded by Aram Galstyan
Content may be subject to copyright.
Opinions described via bump densities (28). Blue curve: the initial opinion of P given by (28) with b = 1. Purple curve: the opinion of Q described by (28) with b ~ 0 : 001 . Olive curve: the resulting opinion of P obtained via (16) with E ~ 0 : 5 . doi:10.1371/journal.pone.0099557.g002
Source publication
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 connec...
Contexts in source publication
Context 1
... if 1 = v Q w 1 = v P (resp. 1 = v Q v 1 = v P ), the final opinion of becomes more (resp. less) narrow than his initial opinion. Fig. 1(b) shows that ( v P e { v P )( v Q { v P ) § 0 holds more generally. Thus, the weighted average approach is a particular case of our model, where the agent P is persuaded by a slightly different opinion. Note also that our model suggests a parameter structure of the weighted average approach. Gaussian densities (with three latitudes) do correspond to the phenomenology of social psychology. However, in certain scenarios one might need other forms of densities, e.g., when the probability is strictly zero outside of a finite support. Such opinions can be represented by bump-functions where b w 0 is a parameter, ( b ) is the normalization and the support of the bump function was chosen to be 1⁄2 { 1,1 for concretness. The advantage of the bump function that is infinitely differentiable despite of having a finite support. For sufficiently large b , x ( x ; b ) is close to a Gaussian, while for small b , x ( x ; b ) represents an opinion that is (nearly) homogeneous on the interval 1⁄2 { 1,1 ; see Fig. 2. The opinion revision with bump densities follows to the general intuition of rule (16); see Fig. 2. One of extensively studied questions in social psychology is how the opinion change is related to the discrepancy between the initial opinion and the position conveyed by the persuasive message [10,35,40,45,69]. Initial studies suggested a linear relationship between discrepancy and the opinion change [35], which agreed with the prediction of the weighted average model. Indeed, (22) yields the following linear relationship between the change in the anchor and the initial opinion discrepancy of P and Q : However, consequent experiments revealed that the linear regime is restricted to small discrepancies only and that the actual behavior of the opinion change as a function of the discrepancy is non-monotonic: the opinion change reaches its maximal value at some discrepancy and decreases afterward [10,40,45,69]. To address this issue within our model, we need to define distance h 1⁄2 p , q between two probability densities p ( x ) and q ( x ). Several such distances are known and standardly employed [32]. Here we select the Hellinger distance (metric) Since p ( x ) is a unit vector in the ‘ 2 norm, Eq. (30) relates to the Euclidean ( ‘ 2 -norm) distance. It is applicable to discrete probabilities by changing the integral in (30, 31) to sum. For Gaussian opinions (17) we obtain A virtue of the Hellinger distance is that it is a measure of overlap between the two densities; see (31). We stress, however, that there are other well-known distances measures in statistics [32]. All results obtained below via the Hellinger distance will be checked with one additional metric, the total variation ( ‘ 1 -norm distance): (To motivate the choice of (33), let us recall two important variational max V [ R 1 D Ð V d features x ( p ( x ) { q ( of x )) D this . (2) Define distance two (generally [32]: (1) dependent) d 1⁄2 p , q ~ that random Ð d x variables g ( x , y ) ~ X q ( , Y y ) , with Ð d y joint g ( x , y probability ) ~ p ( x ) . Now density it g holds ( x , y ) such that d 1⁄2 p , q ~ min 1⁄2 Pr( X ~ 6 Y ) , where Pr( X ~ 6 Y ) ~ 1 { Ð d x g ( x , x ) , and the minimization is taken over all g ( x , y ) with fixed marginals equal to p ( x ) and q ( y ), respectively.) The opinion change is characterized by the Hellinger distance h 1⁄2 p , p between the initial and final opinion of P , while the discrepancy is quantified by the Hellinger distance h 1⁄2 p , q between the initial opinion of P and the persuading opinion. For concreteness we assume that the opinion strengths 1 = v P and 1 = v Q are fixed. Then h 1⁄2 p , q reduces to the distance m ~ D m P { m Q D between the anchors (peaks of p ( x ) and q ( x )); see (32). Fig. 3(a) shows that the change h 1⁄2 p , p is maximal at m ~ m c h ; it decreases for m w m c h , since the densities of P and Q have a smaller overlap. The same behavior is shown by the total variation d 1⁄2 p , p that maximizes at m ~ m c d ; see Fig. 3(a). The dependence of m c h (and of m c d ) on E is also non- monotonic; Fig. 3(b). This is a new prediction of the model. Also, m c h and m c d are located within the latitude of non-commitment of P (this statement does not apply to m c h , when E is close to 1 or 0); cf. (18, 19). This point agrees with experiments [10,69]. Note that experiments in social psychology are typically carried out by asking the subjects to express one preferred opinion under given experimental conditions [10,35,40,45,69]. It is this single opinion that is supposed to change under persuasion. It seems reasonable to relate this single opinion to the maximally probable one (anchor) in the probabilistic set-up. Thus, in addition to calculating distances, we show in Fig. 3(c) how the final anchor m P of P deviates from his initial anchor m P . Fig. 3(c) shows that for E w 0 : 25 , the behavior of D m ~ D m P e { m P D as a function of m ~ D m P { m Q D has an inverted-U shape, as expected. It is seen that D m saturates to zero much faster compared to the distance h 1⁄2 p , p . In other words, the full probability p keeps changing even when the anchor does not show any change; cf. Fig. 3(c) with Fig. 3(a). A curious phenomenon occurs for a sufficiently small E ; see Fig. 3(c) with E ~ 0 : 1 . Here D m drops suddenly to a small value when m passes certain crticial point; Fig. 3(c). The mechanism behind this sudden change is as follows: when the main peak of p ( x ) shifts towards m Q , a second, sub-dominant peak of p ( x ) appears at a value smaller than m P . This second peak grows with m and at some critical value it overcomes the first peak, leading to a bistability region and an abrupt change of D m . The latter arises due to a subtle interplay between the high credibility of Q (as expressed by a relatively small value of E ) and sufficiently large discrepancy between P and Q (as expressed by a relatively large value of m ). Recall, however, that the distance h 1⁄2 p , p calculated via the full probability does not show any abrupt change. The abrupt change of ...
Context 2
... if 1 = v Q w 1 = v P (resp. 1 = v Q v 1 = v P ), the final opinion of becomes more (resp. less) narrow than his initial opinion. Fig. 1(b) shows that ( v P e { v P )( v Q { v P ) § 0 holds more generally. Thus, the weighted average approach is a particular case of our model, where the agent P is persuaded by a slightly different opinion. Note also that our model suggests a parameter structure of the weighted average approach. Gaussian densities (with three latitudes) do correspond to the phenomenology of social psychology. However, in certain scenarios one might need other forms of densities, e.g., when the probability is strictly zero outside of a finite support. Such opinions can be represented by bump-functions where b w 0 is a parameter, ( b ) is the normalization and the support of the bump function was chosen to be 1⁄2 { 1,1 for concretness. The advantage of the bump function that is infinitely differentiable despite of having a finite support. For sufficiently large b , x ( x ; b ) is close to a Gaussian, while for small b , x ( x ; b ) represents an opinion that is (nearly) homogeneous on the interval 1⁄2 { 1,1 ; see Fig. 2. The opinion revision with bump densities follows to the general intuition of rule (16); see Fig. 2. One of extensively studied questions in social psychology is how the opinion change is related to the discrepancy between the initial opinion and the position conveyed by the persuasive message [10,35,40,45,69]. Initial studies suggested a linear relationship between discrepancy and the opinion change [35], which agreed with the prediction of the weighted average model. Indeed, (22) yields the following linear relationship between the change in the anchor and the initial opinion discrepancy of P and Q : However, consequent experiments revealed that the linear regime is restricted to small discrepancies only and that the actual behavior of the opinion change as a function of the discrepancy is non-monotonic: the opinion change reaches its maximal value at some discrepancy and decreases afterward [10,40,45,69]. To address this issue within our model, we need to define distance h 1⁄2 p , q between two probability densities p ( x ) and q ( x ). Several such distances are known and standardly employed [32]. Here we select the Hellinger distance (metric) Since p ( x ) is a unit vector in the ‘ 2 norm, Eq. (30) relates to the Euclidean ( ‘ 2 -norm) distance. It is applicable to discrete probabilities by changing the integral in (30, 31) to sum. For Gaussian opinions (17) we obtain A virtue of the Hellinger distance is that it is a measure of overlap between the two densities; see (31). We stress, however, that there are other well-known distances measures in statistics [32]. All results obtained below via the Hellinger distance will be checked with one additional metric, the total variation ( ‘ 1 -norm distance): (To motivate the choice of (33), let us recall two important variational max V [ R 1 D Ð V d features x ( p ( x ) { q ( of x )) D this . (2) Define distance two (generally [32]: (1) dependent) d 1⁄2 p , q ~ that random Ð d x variables g ( x , y ) ~ X q ( , Y y ) , with Ð d y joint g ( x , y probability ) ~ p ( x ) . Now density it g holds ( x , y ) such that d 1⁄2 p , q ~ min 1⁄2 Pr( X ~ 6 Y ) , where Pr( X ~ 6 Y ) ~ 1 { Ð d x g ( x , x ) , and the minimization is taken over all g ( x , y ) with fixed marginals equal to p ( x ) and q ( y ), respectively.) The opinion change is characterized by the Hellinger distance h 1⁄2 p , p between the initial and final opinion of P , while the discrepancy is quantified by the Hellinger distance h 1⁄2 p , q between the initial opinion of P and the persuading opinion. For concreteness we assume that the opinion strengths 1 = v P and 1 = v Q are fixed. Then h 1⁄2 p , q reduces to the distance m ~ D m P { m Q D between the anchors (peaks of p ( x ) and q ( x )); see (32). Fig. 3(a) shows that the change h 1⁄2 p , p is maximal at m ~ m c h ; it decreases for m w m c h , since the densities of P and Q have a smaller overlap. The same behavior is shown by the total variation d 1⁄2 p , p that maximizes at m ~ m c d ; see Fig. 3(a). The dependence of m c h (and of m c d ) on E is also non- monotonic; Fig. 3(b). This is a new prediction of the model. Also, m c h and m c d are located within the latitude of non-commitment of P (this statement does not apply to m c h , when E is close to 1 or 0); cf. (18, 19). This point agrees with experiments [10,69]. Note that experiments in social psychology are typically carried out by asking the subjects to express one preferred opinion under given experimental conditions [10,35,40,45,69]. It is this single opinion that is supposed to change under persuasion. It seems reasonable to relate this single opinion to the maximally probable one (anchor) in the probabilistic set-up. Thus, in addition to calculating distances, we show in Fig. 3(c) how the final anchor m P of P deviates from his initial anchor m P . Fig. 3(c) shows that for E w 0 : 25 , the behavior of D m ~ D m P e { m P D as a function of m ~ D m P { m Q D has an inverted-U shape, as expected. It is seen that D m saturates to zero much faster compared to the distance h 1⁄2 p , p . In other words, the full probability p keeps changing even when the anchor does not show any change; cf. Fig. 3(c) with Fig. 3(a). A curious phenomenon occurs for a sufficiently small E ; see Fig. 3(c) with E ~ 0 : 1 . Here D m drops suddenly to a small value when m passes certain crticial point; Fig. 3(c). The mechanism behind this sudden change is as follows: when the main peak of p ( x ) shifts towards m Q , a second, sub-dominant peak of p ( x ) appears at a value smaller than m P . This second peak grows with m and at some critical value it overcomes the first peak, leading to a bistability region and an abrupt change of D m . The latter arises due to a subtle interplay between the high credibility of Q (as expressed by a relatively small value of E ) and sufficiently large discrepancy between P and Q (as expressed by a relatively large value of m ). Recall, however, that the distance h ...
Similar publications
This study investigates a new opinion formation model of heterogeneous agents, a network stubborn individuals and orators (NSO) model based on game theory and complex social networks. Game theory solves economists’ rational choice-making problems, and complex social networks reflect the social impact on opinion evolution. The NSO model involves bot...
In this work we study the problem of tax evasion on a fully-connected
population. For this purpose, we consider that the agents may be in three
different states, namely honest tax payers, tax evaders and undecided, that are
individuals in an intermediate class among honests and evaders. Every
individual can change his/her state following a kinetic...
Consider an election with two competing candidates and a set of voters whose opinions change over time. We study the best strategies that can be used by each candidate to influence voters. We also evaluate the knowledge advantage when one of the candidates knows in advance the adversary’s strategy. We prove that an economy of up to 50% of the budge...
This article reports new advancements in the theory of influence system evolution in small deliberative groups, and a novel set of empirical findings on such evolution. The theory elaborates the specification of the single-issue opinion dynamics of such groups, which has been the focus of theory development in the field of opinion dynamics, to incl...
Active motion of living organisms and artificial self-propelling particles has been an area of intense research at the interface of biology, chemistry and physics. Significant progress in understanding these phenomena has been related to the observation that dynamic self-organization in active systems has much in common with ordering in equilibrium...
Citations
... Confirmation bias may have accounted for the different results found when investigating the 2020 presidential election outcome and the 2016 US presidential election outcome (Gibbons et al. 2020). Confirmation bias describes the tendency for individuals to acquire and interpret information in a way that supports their preexisting beliefs (Allahverdyan and Galstyan 2014). Following the announcement declaring Biden the winner of the 2020 US presidential election, many conservatives demonstrated confirmation bias when they engaged in rampant election denialism protesting a "stolen election" with Trump as the "true" victor. ...
Winning and losing voters tend to experience positive and negative emotions toward elections, respectively. The emotions of autobiographical event memories fade over time with unpleasant emotions fading faster than pleasant emotions; this phenomenon is referred to as the fading affect bias (FAB). Although the FAB differs across several event types (e.g., social media and nonsocial media), the FAB and its relations to political, healthy (e.g., grit), and unhealthy variables (e.g., depression) did not differ significantly across political and nonpolitical events in the context of the 2016 US presidential election. To further explore these relations across political and nonpolitical events, the current study used two online samples (college students and MTurk) in the 2020 US presidential election context. The FAB was expected and found to be robust, it was positively predicted by rehearsal ratings, and its relations to political variables were stronger for political events than nonpolitical events.
... Allahverdyan and Galstyan [29] define confirmation bias as the "tendency to acquire or evaluate new information in a manner consistent with one's existing beliefs". Furthermore, Jones and Sugden [30] demonstrated a positive confirmation bias in both information acquisition and use, as evidenced in an experiment where individuals selected "information on what to buy, prior to making a decision". ...
In their process of information seeking, human users tend to selectively ignore information that contradicts their pre-existing beliefs or opinions. These so-called “self-imposed filter bubbles” (SFBs) pose a significant challenge for argumentative conversational agents aiming to facilitate critical, unbiased opinion formation on controversial topics. With the ultimate goal of developing a system that helps users break their self-imposed filter bubbles (SFBs), this paper aims to investigate the role of co-speech gestures, specifically examining how these gestures significantly contribute to achieving this objective. This paper extends current research by examining methods to engage users in cooperative discussions with a virtual human-like agent, encouraging a deep reflection on arguments to disrupt SFBs. Specifically, we investigate the agent’s non-verbal behavior in the form of co-speech gestures. We analyze whether co-speech gestures, depending on the conveyed information, enhance motivation, and thus conversational user engagement, thereby encouraging users to consider information that could potentially disrupt their SFBs. The findings of a laboratory study with 56 participants highlight the importance of non-verbal agent behaviors, such as co-speech gestures, in improving users’ perceptions of the interaction and the conveyed content. This effect is particularly notable when the content aims to challenge the user’s SFB. Therefore, this research offers valuable insights into enhancing user engagement in the design of multimodal interactions with future cooperative argumentative virtual agents.
... In particular, an analogy between mental states and quantum states has been used to explain cognitive dissonance [39] and gender fluidity [20]. A similar approach has been adopted in models of confirmation bias [40,41], gambling [36,42], the prisoner's dilemma game [36,43], conjunction fallacy [36,41], and the Ellsberg paradox [36]. However, although there exist classical models of confirmation bias [40,41], opinion formation and polarisation [44][45][46][47][48][49][50][51][52][53][54][55][56], human morality [57], and backfire effect [15,58], the core principles of quantum mechanics exploited in this paper have not been previously applied to the aforementioned psychological phenomena. ...
... A similar approach has been adopted in models of confirmation bias [40,41], gambling [36,42], the prisoner's dilemma game [36,43], conjunction fallacy [36,41], and the Ellsberg paradox [36]. However, although there exist classical models of confirmation bias [40,41], opinion formation and polarisation [44][45][46][47][48][49][50][51][52][53][54][55][56], human morality [57], and backfire effect [15,58], the core principles of quantum mechanics exploited in this paper have not been previously applied to the aforementioned psychological phenomena. ...
We propose a quantum-mechanical model that represents a human system of beliefs as the quantised energy levels of a physical system. This model represents a novel perspective on opinion dynamics, recreating a broad range of experimental and real-world data that exhibit an asymmetry of opinion radicalisation. In particular, the model demonstrates the phenomena of pronounced conservatism versus mild liberalism when individuals are exposed to opposing views, mirroring recent findings on opinion polarisation via social media exposure. Advancing this model, we establish a robust framework that integrates elements from physics, psychology, behavioural science, decision-making theory, and philosophy. We also emphasise the inherent advantages of the quantum approach over traditional models, suggesting a number of new directions for future research work on quantum-mechanical models of human cognition and decision-making.
... Both a confirmation bias and a primacy effect could explain the stronger influence of early information if the timing of social information also changed the sequence in which individuals deliberated over evidence. Both effects have been described previously in social (e.g., 16,61,62 ) and non-social contexts. ...
People’s decisions are often informed by the choices of others. Evidence accumulation models provide a mechanistic account of how such social information enters the choice process. Previous research taking this approach has suggested two fundamentally different cognitive mechanisms by which people incorporate social information. On the one hand, individuals may update their evidence level instantaneously when observing social information. On the other hand, they may gradually integrate social information over time. These accounts make different predictions on how the timing of social information impacts its influence. The former predicts that timing has no impact on social information uptake. The latter predicts that social information which arrives earlier has a stronger impact because its impact increases over time. We tested both predictions in two studies in which participants first observed a perceptual stimulus. They then entered a deliberation phase in which social information arrived either early or late before reporting their judgment. In Experiment 1, early social information remained visible until the end and was thus displayed for longer than late social information. In Experiment 2, which was preregistered, early and late social information were displayed for an equal duration. In both studies, early social information had a larger impact on individuals’ judgments. Further, an evidence accumulation analysis found that social information integration was best explained by both an immediate update of evidence and continuous integration over time. Because in social systems, timing plays a key role (e.g., propagation of information in social networks), our findings inform theories explaining the temporal evolution of social impact and the emergent social dynamics.
... Owing to the success of ALK-TKIs in ALK + NSCLC, IMTs that share similar traits are believed to exhibit compatible results. Hence, there is a tendency to acquire new data in accordance with previous beliefs [118]. To resolve these issues, larger studies are needed. ...
Opinion statement
Inflammatory myofibroblastic tumor (IMT), characterized by intermediate malignancy and a propensity for recurrence, has presented a formidable clinical challenge in diagnosis and treatment. Its pathological characteristics may resemble other neoplasms or reactive lesions, and the treatment was limited, taking chemotherapies as the only option for those inoperable. However, discovering anaplastic lymphoma kinase (ALK) protein expression in approximately 50% of IMT cases has shed light on a new diagnostic approach and application of targeted therapies. With the previous success of combating ALK⁺ non-small-cell lung cancers with ALK tyrosine kinase inhibitors (TKIs), crizotinib, a first-generation ALK-TKI, was officially approved by the U.S. Food and Drug Administration in 2020, to treat unresectable ALK⁺ IMT. After the approval of crizotinib, other ALK-TKIs, such as ceritinib, alectinib, brigatinib, and lorlatinib, have proven their efficacy on ALK⁺ IMT with sporadic case reports. The sequential treatments of targeted therapies in may provide the insight into the choice of ALK-TKIs in different lines of treatment for unresectable ALK⁺ IMT.
... Classic studies in this field have also revealed that people are more likely to be exposed to information that confirms their existing beliefs, leading to confirmation bias, where individuals selectively perceive and accept information that aligns with their views while dismissing contrary information. [31][32][33] In addition, echo chambers can further generate social environments that reinforce confirmation bias by limiting exposure to diverse perspectives and amplifying existing beliefs. [34][35][36] The interplay between confirmation bias and echo chambers contributes to polarization or disagreement and can hinder open and constructive dialogue between individuals with different opinions or beliefs. ...
Opinion cascades, initiated by active opinions, offer a valuable avenue for exploring the dynamics of consensus and disagreement formation. Nevertheless, the impact of biased perceptions on opinion cascade, arising from the balance between global information and locally accessible information within network neighborhoods, whether intentionally or unintentionally, has received limited attention. In this study, we introduce a threshold model to simulate the opinion cascade process within social networks. Our findings reveal that consensus emerges only when the collective stubbornness of the population falls below a critical threshold. Additionally, as stubbornness decreases, we observe a higher prevalence of first-order and second-order phase transitions between consensus and disagreement. The emergence of disagreement can be attributed to the formation of echo chambers, which are tightly knit communities where agents’ biased perceptions of active opinions are lower than their stubbornness, thus hindering the erosion of active opinions. This research establishes a valuable framework for investigating the relationship between perception bias and opinion formation, providing insights into addressing disagreement in the presence of biased information.
... Third, according to mass media theory, the selective exposure feature refers to agents' tendency to have biases and preferences when receiving messages and opinions [33], [34]. In the DG and FJ models, an agent's weight coefficients towards others' opinions remain fixed and independent of the opinions. ...
According to mass media theory, the dissemination of messages and the evolution of opinions in social networks follow a two-step process. First, opinion leaders receive the message from the message sources, and then they transmit their opinions to normal agents. However, most opinion models only consider the evolution of opinions within a single network, which fails to capture the two-step process accurately. To address this limitation, we propose a unified framework called the Two-Step Model, which analyzes the communication process among message sources, opinion leaders, and normal agents. In this study, we examine the steady-state opinions and stability of the Two-Step Model. Our findings reveal that several factors, such as message distribution, initial opinion, level of stubbornness, and preference coefficient, influence the sample mean and variance of steady-state opinions. Notably, normal agents' opinions tend to be influenced by opinion leaders in the two-step process. We also conduct numerical and social experiments to validate the accuracy of the Two-Step Model, which outperforms other models on average. Our results provide valuable insights into the factors that shape social opinions and can guide the development of effective strategies for opinion guidance in social networks.
... For example, individuals who read an empirical description of exposure therapy for PTSD exhibited greater positive belief change about the treatment than individuals who were assigned to read either a basic or emotional description (Klein et al. 2021). Although some previous research suggests that it is unlikely that people's opinions will change solely because of facts due to personal biases (e.g., Allahverdyan, Galstyan, and Branas-Garza 2014), past research on didactic psychoeducation to promote psychotherapy engagement among BIPOC individuals have shown promising results (Alvidrez, Areán, and Stewart 2005;Ghafoori et al. 2016;Shin and Lukens 2002). ...
Psychedelic-assisted psychotherapy (PAP) is gaining renewed interest as a treatment for various mental disorders. However, there has been limited Black, Indigenous, and People of Color (BIPOC) representation in PAP clinical trials, signaling the need for culturally consonant communication about the efficacy and safety of PAP. We randomly assigned 321 BIPOC and 301 non-Hispanic White participants to four different modes of psychoeducation (didactic, visual, narrative, hope-based) and tested effects on likelihood of seeking and referring others to PAP using ANCOVAs. The influences of different psychoeducation components on these likelihoods were also tested using hierarchical regression modeling. Regardless of psychoeducation mode, BIPOC participants were more likely to seek PAP than non-Hispanic White participants after psychoeducation. Further, information on physical safety and success rate of PAP uniquely predicted BIPOC participants' likelihood of seeking and referring others to PAP after psychoeducation. Our findings suggest that once provided psychoeducation, BIPOC participants are receptive to seeking or referring others to PAP. BIPOC participants also appear to prioritize physical safety and rate of success of PAP in these decisions. Stigma against PAP is likely not the primary barrier to recruitment of BIPOC individuals into PAP trials. Instead, researchers should conduct more psychoeducational outreach to diversify future trials.
... In particular, an analogy between mental states and quantum states has been used to explain cognitive dissonance (29). A similar approach has been adopted in models of confirmation bias (30,31), two-stage gambling (28), Prisoner's Dilemma game (28), conjunction fallacy (28,31) and Ellsberg paradox (28). However, although there exist classical models of confirmation bias (30,31), opinion formation and polarization (32)(33)(34)(35)(36)(37), and backfire effect (38,39), the principles of quantum mechanics have not been applied to these psychological phenomena yet. ...
... A similar approach has been adopted in models of confirmation bias (30,31), two-stage gambling (28), Prisoner's Dilemma game (28), conjunction fallacy (28,31) and Ellsberg paradox (28). However, although there exist classical models of confirmation bias (30,31), opinion formation and polarization (32)(33)(34)(35)(36)(37), and backfire effect (38,39), the principles of quantum mechanics have not been applied to these psychological phenomena yet. ...
We propose a quantum-mechanical model that represents a human system of beliefs as quantized energy levels of a physical system. This model underscores a novel perspective on opinion dynamics, recreating a broad range of experimental and real-world data that exhibit an asymmetry of opinion radicalization. In particular, the model demonstrates the phenomena of pronounced conservatism versus mild liberalism when individuals are exposed to opposing views, mirroring recent findings on opinion polarization via social media exposure. Advancing this model, we establish a solid framework that integrates elements from physics, psychology, and philosophy, and also emphasize the inherent advantages of the quantum approach over traditional classical models. Our findings enhance ongoing efforts to elucidate psychological, behavioral, and decision-making phenomena using the laws of quantum mechanics. These results also lend additional support to the quantum mind hypothesis, suggesting that the most radicalized opinions may not necessarily mirror the most firmly held beliefs but are driven by strongly held opinions and exposure to dissenting views.
... At the same time, Becker et al. [18] show network conditions where social influence may improve the accuracy of group estimates, despite the reduction in the diversity of opinions. Other classes of models seek to incorporate cognitive biases in the agents' information update process, e.g., explicitly modelling confirmation bias by having agents place more weight on opinions that already conform to their existing beliefs [19,20,21]. Bounded confidence models instead impose that agents only interact if their opinions are similar enough [22,21]. ...
We develop a model of opinion dynamics where agents in a social network seek to learn a ground truth among a set of competing hypotheses. Agents in the network form private beliefs about such hypotheses by aggregating their neighbors' publicly stated beliefs, in an iterative fashion. This process allows us to keep track of scenarios where private and public beliefs align, leading to population-wide consensus on the ground truth, as well as scenarios where the two sets of beliefs fail to converge. The latter scenario - which is reminiscent of the phenomenon of cognitive dissonance - is induced by injecting 'conspirators' in the network, i.e., agents who actively spread disinformation by not communicating accurately their private beliefs. We show that the agents' cognitive dissonance non-trivially reaches its peak when conspirators are a relatively small minority of the population, and that such an effect can be mitigated - although not erased - by the presence of 'debunker' agents in the network.
