Certain members of society, who are characterized by overconfidence, a relatively low level of knowledge, and a high level of misinformation or incomplete information, tend to selectively choose information from real life or the digital environment, in order to confirm their earlier incorrect beliefs. This phenomenon, known as "confirmation bias", is widely recognized in every society and mainly occurs in political discussions (by self-proclaimed "political analysts"), but also in discussions on the topic of global conspiracy theories. This kind of phenomenon is increasingly present in digital communication among consumers. Such individuals unfoundedly influence their followers (via electronic word of mouth or writing inaccurate reviews) by favoring one product or service over competing ones with baseless, truncated information or misinformation. Spreading misinformation in the digital environment can have very dangerous consequences. There are a large number of practical examples in which the spread of misinformation led to riots in the streets and violent behavior that even caused deaths.

It is well established that people tend to favor confirming strategies in hypothesis testing. However, little is known about individual differences in this topic. The present study has two aims: (a) adapting existing experimental tasks on hypothesis testing to measure individual differences, (b) investigating the convergent and discriminant properties of these measures within a multitrait–multimethod framework using a confirmatory factor analysis approach. Participants (N = 200) completed a total of nine behavioral tasks: three tasks (two rule hypothesis testing tasks and a trait hypothesis testing task) measuring three aspects of confirmation bias in information processing (information search, weighing of evidence, memory recall). Scores to most tasks reached satisfactory levels of internal consistency, showing that individual differences in confirmation bias can be reliably measured. The multitrait–multimethod analyses provided moderate evidence for the convergent validity and small to moderate evidence for the discriminant validity of the measures. These findings reveal that confirmation bias is a more unified phenomenon than what has been suggested. Confirmation bias in hypothesis testing 2 Abstract It is well established that people tend to favor confirming strategies in hypothesis testing. However, little is known about individual differences in this topic. The present study has two aims: (a) adapting existing experimental paradigms on hypothesis testing to measure individual differences in confirmation bias, and (b) investigating the convergent and discriminant validity of these measures within a multitrait–multimethod framework using a confirmatory factor analysis approach. Participants (N = 200) completed a total of nine behavioral tasks: three paradigms (two rule hypothesis testing paradigms and a trait hypothesis testing paradigm) measuring three components of confirmation bias in information processing (information search, weighing of evidence, memory recall). Scores to most tasks reached satisfactory levels of internal consistency, showing that individual differences in confirmation bias can be reliably measured. The multitrait–multimethod analyses provided moderate evidence for the convergent validity and small to moderate evidence for the discriminant validity of the measures. These findings reveal that confirmation bias is a more unified phenomenon than what has been suggested. Keywords: Confirmation bias; Hypothesis testing; Individual differences; Measurement Confirmation bias in hypothesis testing 3 “I had also, during many years, followed a golden rule, namely, that whenever a published fact, a new observation or thought came across me, which was opposed to my general results, to make a memorandum of it without fail and at once; for I had found by experience that such facts and thoughts were far more apt to escape from the memory than favourable ones. Owing to this habit, very few objections were raised against my views which I had not at least noticed and attempted to answer.” Charles Darwin, The Autobiography of Charles Darwin 1. Introduction Confirmation bias is commonly defined as the fact that “information is searched, interpreted, and remembered in such a way that it systematically impedes the possibility that the hypothesis will be rejected” (Oswald & Grosjean, 2004, p.79). Despite the apparent simplicity of the phenomenon, this definition already implies that confirmation bias is an umbrella term that encompasses at least three related though distinct stages (seeking, interpreting or remembering) of information processing (Nickerson, 1998; Klayman, 1995; Oswald & Grosjean, 2004). Yet, and although confirmation bias has been described as “the best known and most widely accepted notion of inferential error to come out of the literature on human reasoning” (Evans, 1989, p. 41), it is still unclear how these three different components of confirmation bias are related to each other. The goal of the present study is to address this question, by examining individual differences. Before we do so, however, one should distinguish further between two aspects of confirmation bias commonly studied in the literature, the motivational aspect and the more cognitive aspect, as the present study will only focus on the latter. On the one hand, the motivational aspect is the tendency to support one’s own prior views (Lord et al.,1979), sometimes referred to as a myside bias (Barron, 1995; Mercier 2017). On the other hand, the cognitive aspect of confirmation bias concerns hypotheses that are not specifically related to the beliefs or values of the individual (Baron, 2008; Poletiek, 2001), as for instance in the pioneer work of Wason (1960, 1966), which we shall describe now. Confirmation bias in hypothesis testing 19 correlations across components) should have been found. Rather, information search, weighing of evidence, and memory recall shared a substantial amount of common variance indicating the presence of a general latent factor that captures individual differences in the tendency to confirm the hypothesis in hand. Future studies should further clarify the construct validity of confirmation bias, by relying upon the tasks introduced in this study and relating the scores to external criteria (e.g., real-life outcomes). Future research should also conduct a CFA–MTMM analysis on myside bias to test if our results hold for the motivational aspect of confirmation bias. For that purpose, one might build on the unified paradigm designed by Vedejová and Čavojová (2022). Another important avenue of research will be to investigate the correlations between the cognitive and the motivational components of confirmation bias. This research would allow one to test further whether this ubiquitous phenomenon is actually even more unified than what has been previously suggested.

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 9 7 2014 9 7 e99557 e99557 14 7 2014 © 2014 Allahverdyan, Galstyan This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. 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. 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. status released display-pdf yes is-olf no is-manuscript no is-preprint no is-journal-matter no is-scanned no is-retracted no Received 2013 Dec 8; Accepted 2014 May 15; Collection date 2014. Introduction Confirmation bias is the tendency to acquire or process new information in a way that confirms one's preconceptions and avoids contradiction with prior beliefs [52] . Various manifestations of this bias have been reported in cognitive psychology [5] , [67] , social psychology [24] , [54] , politics [46] and (media) economics [31] , [51] , [57] , [73] . Recent evidence suggests that scientific practices too are susceptible to various forms of confirmation bias [12] , [38] , [43] , [44] , [52] , even though the imperative of avoiding precisely this bias is frequently presented as one of the pillars of the scientific method. Here we are interested in the opinion revision of an agent who is persuaded (or advised) by another agent [10] , [13] , [52] . We also demonstrate that the rule reproduces several basic features of the cognitive dissonance phenomenon and predicts new scenarios of its emergence. Finally, the so called boomerang ( backfire ) effect can emerge as an extreme form of confirmation bias. The effect is given a straightforward mathematical description in qualitative agreement with experiments. The rest of this paper is organized as follows. In the next section we introduce the problem setup and provide a brief survey of relevant work, specifically focusing on inadequacy of the standard Bayesian approach to opinion revision under persuasion. However, that modification betrays the spirit of the normative Bayesianism, because it makes conditional probabilities depending on the prior probability.) It is worthwhile to note that researchers have studied several aspects of confirmation bias by looking at certain deviations from the Bayes rule, e.g. when the conditional probability are available, but the agent does not apply the proper Bayes rule deviating from it in certain aspects [31] , [51] , [57] , [73] . One example of this is when the (functional) form of the conditional probability is changed depending on the evidence received or on the prior probabilities. Another example is when the agent does not employ the full available evidence and selects only the evidence that can potentially confirm his prior expectations [39] , [48] , [67] . More generally, one has to differentiate between two aspects of the confirmation bias that can be displayed either with respect to information acquiring, or information assimilation (or both) [52] . Our study will concentrate on information assimilation aspect; first, because this aspect is not studied sufficiently well, and second, because because it seems to be more directly linked to cognitive limitations [52] . (Note that the order of presentation effect is not predicted by the Bayesian approach; see (2).) Some studies suggest that the first opinion matters more (primacy effect), whereas other studies advocate that the last interaction is more important (recency effect). While it is not completely clear which experimentally (un)controlled factors are responsible for primacy and recency, there is a widespread tendency of relating the primacy effect to confirmation bias [13] , [52] . This relation involves a qualitative argument that we scrutinize below. We now define the order of presentation effect in our situation. only in the special case). Hence, in future it needs to be extended by involving additional mechanisms, e.g. those related to “long-term memory” processes which could be responsible for the above experimental fact. Recall in this context there are several other theoretical approaches that address the primacy-recency difference [11] , [34] , [42] , [56] , [66] . (ii) The model can be used to describe the phenomenon of cognitive dissonance and to formalize the main scenario of its emergence. (iii) Repeated persuasions display several features implying monotonous change of the target opinion towards the persuading

Research on the phenomenon of selective exposure to information demonstrates that after preliminary or final decisions, people show a preference for supporting rather than conflicting information (confirmation bias). In this article, we examine conditions that increase or decrease distortions in the search for information. We report on four experiments indicating that the confirmation bias is influenced by whether people focus on their decision or on the presented pieces of information during the information search. Focusing on the decision, for example, because a reward for a correct decision