---
title: "Reconciling scientific and commonsense values to improve reasoning"
citation: "Cusimano, C., & Lombrozo, T. (2021). Trends in Cognitive Sciences, 25, 937-949."
year: 2021
doi: "10.1016/j.tics.2021.06.004"
source_pdf: "https://coreycusimano.net/docs/cusimano_lombrozo_2021_tics.pdf"
extraction_note: "Machine-readable Markdown auto-extracted from the published PDF via Docling (text + tables); figures exported separately."
---

# Reconciling scientific and commonsense values to improve reasoning

Cusimano, C., & Lombrozo, T. (2021). Trends in Cognitive Sciences, 25, 937-949. <https://doi.org/10.1016/j.tics.2021.06.004>

**Plain-language summary (author):** Coordinates cognitive models of biased reasoning, normative theories of belief formation, and the lay ethics of belief to identify new ways to improve everyday reasoning.

[Full-text PDF](https://coreycusimano.net/docs/cusimano_lombrozo_2021_tics.pdf)

---

## Trends in Cognitive Sciences

## Review

## Reconciling scienti fi c and commonsense values to improve reasoning

Corey Cusimano 1, * and Tania Lombrozo 1

Scienti fi c reasoning is characterized by commitments to evidence and objectivity. New research suggests that under some conditions, people are prone to reject these commitments, and instead sanction motivated reasoning and bias. Moreover, people's tendency to devalue scienti fi c reasoning likely explains the emergence and persistence of many biased beliefs. However, recent work in epistemology has identi fi ed ways in which bias might be legitimately incorporated into belief formation. Researchers can leverage these insights to evaluate when commonsense af fi rmation of bias is justi fi ed and when it is unjusti fi ed and therefore a good target for intervention. Making reasoning more scienti fi c may require more than merely teaching people what constitutes scienti fi c reasoning; it may require af fi rming the value of such reasoning in the fi rst place.

## The relationship between normative and everyday standards of reasoning

Many philosophers of science and epistemologists argue that, when forming and evaluating beliefs about the world, one ought to be objective, open-minded, and bound by evidence [1-4]. This kind of reasoning has been canonized as scientific reasoning (see Glossary), with norms of objectivity and evidentialism identi fi ed as de fi ning norms of the scientific ethos [5,6]. According to the scienti fi c ethos, inquirers can and should incorporate their preferences and values into some aspects of inquiry (such as determining which questions are worth answering; Box 1), but the process by which they form beliefs about the world ought to be objective and based purely on evidence. And yet, everyday belief formation often violates these norms and is instead thoroughly infused with bias, motive, and value. This observation raises two pressing questions. First, do people actually agree that violating norms of objectivity and evidentialism is bad? In other words, does the lay ethics of belief share the commitments of the scienti fi c ethos? And second, what do people's beliefs about what constitutes good reasoning entail for scientists who aspire to improve reasoning?

A long-standing assumption in cognitive science has been that the lay ethics of belief approximates the scienti fi c ethos, such that failures to live up to objectivity or evidentialism re fl ect unconscious biases [7-9]. However, recent studies have called this assumption into question. In many contexts, people explicitly devalue scienti fi c norms of reasoning and instead regard biased reasoning as good reasoning. For example, people sometimes believe it is appropriate to assume a friend's innocence (even when the evidence suggests otherwise), or inappropriate to judge an individual based on the statistics of their social group [10]. Such departures from objectivity fi nd counterparts in recent arguments within epistemology, where philosophers have defended alternative norms to those espoused by the scienti fi c ethos [11]. These new developments raise important questions about when and how psychologists should try to align the lay ethics of belief with the scienti fi c ethos. We suggest that insights from epistemology can inform psychological inquiry by helping researchers identify and evaluate widespread intuitions about what constitutes good reasoning. This can in turn provide a promising strategy for changing everyday standards for reasoning by correcting any misconceptions that underlie them. We propose that this coordination between

## Highlights

People espouse a ' lay ethics of belief ' that de fi nes standards for how beliefs should be evaluated and formed.

People vary in the extent to which they endorse scienti fi c norms of reasoning, such as evidentialism and impartiality, in their own norms of belief. In some cases, people sanction motivated or biased thinking.

Variation in endorsement of scienti fi c norms predicts belief accuracy, suggesting that interventions that target norms could lead to more accurate beliefs.

Normative theories in epistemology vary in whether, and how, they regard reasoning and belief formation as legitimately impacted by moral or pragmatic considerations.

Psychologists can leverage knowledge of people's lay ethics of belief, and normative arguments about when and whether bias is appropriate, to develop interventions to improve reasoning that are both ethical and effective.

1 Department of Psychology, Princeton University, Princeton, NJ, 08540, USA

*Correspondence: cusimano@princeton.edu (C. Cusimano).

## Box 1. Roles for values in scienti fi c inquiry

A major project in the philosophy of science and epistemology has been to characterize how moral and social values should (and should not) in fl uence inquiry and belief [85,86]. Philosophers agree that there are several ways social values ought to in fl uence inquiry (Figure I). First, the questions we ask ought to re fl ect the observation that not all knowledge is equally valuable. This is why scientists articulate ' broader impacts' when submitting grants. Second, the value of acquiring particular knowledge can be outweighed by moral and practical costs. This is why experiments involving humans and other animals must be evaluated by an Institutional Review Board. And third, the effort and resources spent on inquiry should re fl ect the costs and bene fi ts associated with getting things right or wrong. When making risky or important decisions, for example, scientists and decision makers are held to high standards of evidence, which is sometimes codi fi ed as standards of ' due diligence ' or as evidential ' readiness levels ' (e.g., [87]).

There is striking continuity between normative and lay attitudes concerning these indirect roles for values in inquiry. Students put more effort into learning about topics that they think are important [88-90] and people are more curious to learn about topics that they expect to be useful [91-93]. Additionally, people consider the costs of inquiry and knowledge acquisition against other costs. For instance, if additional evidence gathering will be especially costly, say because people will miss an opportunity to make a decision, then they will suspend evidence gathering [8]. Likewise, people spend more time collecting evidence, and more energy scrutinizing that evidence, for decisions involving signi fi cant risk [94,95].

While there is broad expert and intuitive agreement that values should in fl uence these aspects of inquiry, normative claims and descriptive practice diverge when it comes to the role of values in forming beliefs . Many scholars believe that values should not affect the factual beliefs that result from inquiry. For instance, values might in fl uence the decision to pursue research to produce a vaccine and even the level of evidence demanded before the vaccine is distributed within and beyond clinical trials, but values should not in fl uence whether scientists believe that the vaccine is safe or effective. In the main text we argue that this is where the scienti fi c ethos and the lay ethics of belief diverge: For many people, moral and personal concerns are perceived to have a legitimate role not only in inquiry, but also in forming beliefs.

Figure I. The relationship between value, inquiry, and belief in the scienti fi c ethos.

normative epistemology, the lay ethics of belief, and psychological models of biased reasoning can shape educative interventions that have the potential to be both effective and normatively defensible.

## Are people dedicated (but mediocre) disciples of the scienti fi c ethos?

At fi rst glance, the lay ethics of belief appears to resemble the scienti fi c ethos. Across the lifespan and across cultures, people display an af fi nity for acquiring and sharing knowledge [12]. Consistent with a motivation to learn, most people report that they want their beliefs to be based on evidence and sound reasoning [13]. Indeed, the dominant view in social and cognitive psychology is that people not only

## Glossary

## Actively open-minded thinking (AOT)

scale: scale measuring lay attitudes about the importance of considering alternatives to one's current beliefs, changing one's mind in response to new evidence, and avoiding dogmatic or absolutist thinking. This scale closely resembles a disposition towards what we label scienti fi c reasoning.

Biased belief: a belief that is the product of biased reasoning.

Biased reasoning: belief formation characterized by a goal to adopt or avoid adopting a particular belief, or otherwise, a belief formed on the basis of evidence that has itself been weighted or interpreted in light of one's values; eschews at least one of evidentialism or objectivity.

Evidentialism: a normative standard of belief according to which someone only ought to hold beliefs that are based on suf fi cient evidence.

Lay ethics of belief: norms that guide everyday belief evaluation, which may or may not accord with the scienti fi c ethos.

Objectivity: a norm of reasoning according to which evidence is features of the evidence that are

evaluated without regard for values or irrelevant to truth.

Scientific ethos: a normative model of belief formation stating that scienti fi c reasoning is the only permissible and justi fi ed form of belief formation.

Scientific reasoning: belief formation characterized by a goal to be accurate and so also characterized by evidentialism, objectivity, and openmindedness. This can include the subset of motivated reasoning that is non-directional and motivated by accuracy.

Unbiased belief: a belief that is the product of scienti fi c reasoning.

## Trends in Cognitive Sciences

endorse the scienti fi c ethos, but naively believe that they are its paragons [7,9,14]. That is, many psychologists attribute to lay reasoners the meta-beliefs that they reason scienti fi cally and that anyone else reasoning as they do, and with access to the same evidence, would share the same beliefs [9,14].

The assumptions that people want to be scienti fi c reasoners, think that they are, and demand that others be so as well have in fl uenced how psychologists explain the existence and persistence of biased reasoning. Explanations for biased reasoning are manifold, but many share the assumption that biased reasoning is unintentional and that the psychological mechanisms that generate biased beliefs are unconscious. That is, people are unaware that they search for evidence in biased ways [15], unaware that they apply pernicious double-standards for preferenceconsistent and preference-inconsistent propositions [16], unaware that they over-rely on discredited sources of information (such as intuition [17]), unaware that they under-rely on useful information (such as base rates [18]), and so on. Even when people are motivated to hold beliefs that con fl ict with their evidence (and these motivations impact what they believe), it is widely thought that they deny any in fl uence [19]. In relegating violations of the scienti fi c ethos to unconscious processes, these views explain how people can maintain a subjective sense that their beliefs, even the highly biased ones, are consistent with their commitment to the scienti fi c ethos.

Based on this traditional portrait of biased reasoning, a sensible strategy for improving reasoning is to educate people about how to better live up to their own ideals [1,20]. Some such interventions have been successful [21]; however, teaching people how to reason rarely generalizes beyond the training context or domain [22-24]. This is a puzzling result if people genuinely aspire to satisfy the norms of reasoning they are being taught to uphold. Of course, transfer learning is hard, especially when the target of intervention is a domain-general skill [25,26]. But we suggest that an additional factor is at work: in many areas of everyday life, the traditional view that people's ethic of belief aligns with the scienti fi c ethos is wrong.

People may not, in fact, think that they should always approach questions about what to believe ' scienti fi cally ' . When evaluating their own and others' beliefs, people sometimes reject the principles of objectivity and evidentialism and instead maintain that considerations of what is morally or socially good ought to in fl uence what they believe. This suggestion complements prior proposals that people sometimes depart from the mold of ' intuitive scientist ' , instead acting as intuitive politicians, theologians, or prosecutors, and in so doing, knowingly adopt belief goals that eschew accuracy [27]. It also echoes prior critiques in the decision-making literature that putatively irrational behaviors in fact re fl ect people's sensitivity to different (perhaps defensible) normative standards [28,29]. Yet, it is only very recently that evidence has emerged to support the claim that many people explicitly recognize and endorse nonscienti fi c norms for belief and that these norms support and maintain biased beliefs.

## The bounds of scienti fi c reasoning in the lay ethics of belief

If people internalize the scienti fi c ethos, then they should dutifully recommend that others adopt unbiased beliefs and they should evaluate biased beliefs as unjusti fi ed and impermissible. However, recent studies show that, beyond considerations of evidence or objectivity, people evaluate beliefs on moral and pragmatic grounds. As a result, people sometimes prescribe biased beliefs to others.

In the fi rst studies to suggest that people prescribe biased beliefs, participants read about characters facing a decision about what to believe, such as judging whether they would win an essay contest or succeed in a business venture [30]. Participants indicated whether the character should be accurate by selecting a value on a nine-point response scale with ' accurate ' as the midpoint and ' extremely pessimistic ' and ' extremely optimistic ' as the lowest and highest anchors, respectively.

Participants tended to prescribe optimism over accuracy. However, it is unclear whether optimism wasassociated with an inaccurate belief or, instead, with a feeling or attitude, along the lines of ' staying positive ' [31]. Recent work has replicated prescribed departures from accuracy while avoiding this ambiguity and has additionally identi fi ed the grounds on which people recommend bias.

One recent study found that people think overcon fi dence can be detrimental or bene fi cial, depending on the context, and that they prescribe beliefs that depart from accuracy on this basis [32]. For instance, overcon fi dence should be detrimental to decision makers when deliberating about what to do, since it could lead to poor decisions. However, people also tend to think that con fi dence can be motivating and so may prescribe overcon fi dence to people who have already committed to a decision and need motivation to follow through. If people prescribe beliefs to others based in part on how bene fi cial they judge those beliefs to be, rather than just how evidentially sound they are, then they ought to prescribe overcon fi dence to others when aiming to motivate them, but not when informing decision making. This is precisely what has been found: when someone is deciding whether to open a small business, participants think that person should be clear-eyed about their chance of success. But when someone has already started their business, participants think that person should be overcon fi dent.

Additional evidence that a belief's anticipated consequences can factor into its perceived justi fi cation comes from research on scienti fi c and religious believers [33]. In these studies, most participants agreed that expert testimony and strong evidence are good grounds for holding a belief, consistent with the scienti fi c ethos. However, some justi fi cations suggested that participants accepted nonscienti fi c reasons as good grounds for belief, too. For instance, many participants thought that ' leading to ethical behavior ' constitutes a good reason for belief. This fi nding, in conjunction with the results already reviewed, shows that beliefs are evaluated in part on the basis of their consequences, such as helping people behave successfully or ethically, rather than their evidential value alone.

The most direct examination of the lay ethics of belief comes from experiments investigating people's reactions to situations that pose a dilemma between believing based on evidence and believing based on what is morally laudatory [10]. For instance, in one study, participants read about someone who had evidence that their spouse is terminally ill (a doctor informs them that there is a low chance of recovery) but who also knows that staying optimistic will improve their spouse's emotional well-being in the months to come. Participants reported what that person would believe if they formed a belief based solely on their evidence and also reported what that person ' ought ' to believe. If participants endorsed norms of scienti fi c reasoning, these two judgments should be identical. Instead, participants reported that others should hold a more optimistic belief than warranted by their evidence and did so to the extent they thought over-optimism improved others' welfare, signaled loyalty (e.g., giving a friend the bene fi t of the doubt), or indicated respect (e.g., treating someone as an individual). Strikingly, participants sometimes reported that someone was more justi fi ed in holding an evidentially unsupported (but morally bene fi cial) belief than in holding an evidentially supported (but morally risky) belief. For instance, in another vignette, participants judged that a newlywed was more justi fi ed to believe that he had a 0% chance of divorce than a high chance of divorce, even though the same participants reported that a high chance of divorce better re fl ected the newlywed's own evidence (Figure 1). Thus, in the lay ethics of belief, moral considerations do not just in fl uence how people evaluate others' beliefs, they sometimes dominate these evaluations.

Taken together, these studies provide strong evidence against the view that the lay ethics of belief accords with the scienti fi c ethos. People evaluate biased reasoning as legitimate when it can be self-ful fi lling, improve someone's behavior, or further moral ends. As we review next, widespread

## Trends in Cognitive Sciences

## [Trends in Cognitive Sciences](Image of Figure 1)

Figure 1. Moral obligations sometimes override objectivity during third-party belief evaluation. Case study: evaluations of others' beliefs about divorce [10]. (A) Most participants reported that a groom had an obligation to remain optimistic about marriage and, simultaneously, reported that the groom's evidence favored divorce. (B) Participants then read that the groom either formed the morally sanctioned belief (0% chance of divorce) or the evidence-based belief (70% chance of divorce). (C) Participants then evaluated the groom's belief. Participants judged that the morally good (but biased) belief was morally better and more justi fi ed. Moreover, participants more strongly agreed that the groom had suf fi cient evidence for his belief when it was biased versus objective (see Box 2 for more detail regarding this fi nding).

endorsement of nonscienti fi c norms helps explain the prevalence of biased reasoning and biased belief. Contra the view that biased reasoning is unconscious or maligned, biased reasoning is sometimes intended, encouraged, and enforced.

## Devaluing the scienti fi c ethos may explain common and persistent misbelief

## Intrapersonal processes

Nonscienti fi c norms for belief could explain how people come to hold or maintain biased beliefs by either shaping their reasoning process or by affecting their attitudes towards particular beliefs they already hold. Illustrating the former, people could adopt belief-formation practices that often support biased reasoning at an unconscious level (such as biasing their exposure to evidence or holding beliefs to biased standards [7,34-36]), but with full conscious awareness and endorsement. Illustrating the latter, people could feel justi fi ed in maintaining beliefs that they regard as bene fi cial, or in rejecting those they regard as detrimental, and so resist demands from the scienti fi c ethos to scrutinize them. If people's norms for belief have such effects, then variation in norms should be associated with variation in belief.

Consistent with this prediction, recent studies show that variation in peoples' tendency to af fi rm scienti fi c norms of reasoning predicts how people tend to reason and how accurate their beliefs tend to be as a result. For instance, more strongly endorsing ideals of logical consistency and evidentialism is negatively associated with religious, supernatural, and paranormal belief [13,37-40]. Similarly, individuals who score high on the actively open-minded thinking (AOT) scale (and so, for instance, more strongly endorse statements like ' people should take

## Trends in Cognitive Sciences

into consideration evidence that goes against their beliefs ' ) are more likely to hold beliefs supported by evidence, such as believing in anthropogenic climate change [38,41,42]. AOT is also associated with more accurate evaluations of arguments [43] and lower susceptibility to cognitive biases [44,45] and it predicts a greater tendency to collect and integrate additional information into one's beliefs [46]. As a consequence, people who value such thinking make more accurate predictions, from basketball wins [46] to major world events [47].

Scienti fi c norms of belief may also protect individuals against others' biased reasoning. People high in AOT temper their con fi dence in others who engage in biased reasoning, such as ineptly weighing the balance of evidence [48]. And in studies on social media behavior, these individuals make more accurate assessments of others [49] and are less susceptible to fake news [50]. These fi ndings reveal a systematic association between the value placed on scienti fi c norms of belief, on the one hand, and the types of beliefs that individuals ultimately endorse, on the other.

## Interpersonal processes

Norms for belief are also likely to perpetuate bias and poor reasoning by affecting people's evaluations of others. People often think others have control over what they believe and, accordingly, hold them responsible and blameworthy when they hold undesirable beliefs [51,52]. In turn, people try to form (or at least report) beliefs that they think others want them to hold [53,54]. Thus, norms for belief have consequences for the kinds of reasoning and corresponding beliefs that will be rewarded and in turn what kinds of beliefs people are motivated to form. People who value scienti fi c reasoning judge others harshly for being illogical or ignoring evidence [13]. In contrast, people who tend to value a belief's moral qualities think it is impermissible for others to believe on the basis of evidence at the expense of morality [10].

Social transmission of biased reasoning need not occur intentionally through explicit social approval or disapprobation. People align their beliefs with what they think those around them believe and pay particular attention to prestigious and con fi dent members of their community [55,56]. For instance, displays of overcon fi dence from one person can cause observers to reason in ways that lead to overcon fi dence in themselves, a process that can then extend to other members of the community [57]. Likewise, people copy the poor evidential standards of their peers, thereby propagating inaccurate beliefs through groups [58]. Norms that favor biased reasoning can propagate given these dynamics of social transmission: people are more likely to feel (and so act) overcon fi dent if they reason in accordance with norms that eschew accuracy; and people with biased beliefs are more likely to acquire positions of prestige when norms of belief reward bias and devalue objectivity.

Just as endorsing the scienti fi c ethos in fl uences the beliefs of individuals, it is likely that groups that endorse and act on scienti fi c norms of reasoning will end up with more accurate beliefs. Like individuals, groups that engage in comprehensive evidence gathering and scienti fi c reasoning make better decisions [59]. And indeed, the recent renaissance in psychological science speaks to the power of the scienti fi c ethos in group settings: increased emphasis on transparency and higher standards of evidence in psychological science have led to the identi fi cation and correction of widely held falsehoods in the scienti fi c canon and curtailed the dissemination of new errors [60].

## Evaluating and intervening on the lay ethics of belief

Though many scholars endorse the scienti fi c ethos, a survey of the normative arguments favoring a pure and steadfast dedication to objectivity and evidentialism reveals that each argument faces formidable objections (Table 1). For instance, some have questioned whether these values ought to outweigh norms of friendship and loyalty [61]. If they should not, then the common practice of forming partial and overly favorable beliefs about one's relationships [62] may be justi fi ed.

## Trends in Cognitive Sciences

Table 1. Normative arguments for scienti fi c reasoning (or against biased reasoning) and common objections

| Appeals to … | Typical line of reasoning | Typical objection | Refs |
|----------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-----------------|
| The instrumental value of scienti fi c reasoning | More complete and more accurate representations of the world enable more optimal behavior and, therefore, better outcomes. | If the reason to reason scienti fi cally is the moral or pragmatic bene fi t of doing so, then people should depart from scienti fi c reasoning if it would be morally or pragmatically valuable to do so. | [1,24,70-72] |
| The intrinsic value of scienti fi c reasoning | Pursuing truth and knowledge are inherently valuable activities. Engaging in scienti fi c reasoning is an intellectual virtue and a mark of virtuous character. | Some beliefs re fl ect values related to bene fi cence, respect, or loyalty. Even if pursuing truth is intrinsically valuable, it is not clear why these values ought to outweigh moral values when they con fl ict. | [2,61,73,74] |
| The function of belief | Beliefs, by de fi nition, are supposed to accurately represent the world. Therefore, beliefs are only properly evaluated with respect to accuracy-preserving norms. | Grounding the ethics of belief in the function of belief leaves open why people ought to desire well-functioning beliefs. | [4,75] |
| The dif fi culty of purposefully adopting biased beliefs | Beliefs are constrained by evidence. It is unfair to demand that people form biased beliefs if they are incapable of doing so. | It is controversial to what degree beliefs are constrained by evidence. If people can regulate their beliefs, then they plausibly have moral and practical reasons to exercise that capacity. | [8,51,52,68,76] |

Likewise, some scholars have argued that forming beliefs about others based on statistical demographic information denies those individuals due respect [63-65]. And some scholars have argued that belief in the existence of powerful and benevolent deities, free will, or karma [66,67] is justi fi ed on the basis of the practical bene fi ts those beliefs confer to their adherents [68,69]. If these arguments are successful, then biased reasoning may not be bad reasoning and norms that may be ideal in the domain of science may not be ideal in the domain of everyday life. In the sections that follow, we draw connections between these arguments, the lay ethics of belief, and everyday motivated reasoning. We then articulate a strategy for modifying people's valuation of scienti fi c reasoning.

## The relationship between biased reasoning, normative theory, and commonsense values

Philosophers who have argued that values play a legitimate role in belief formation have suggested two ways in which this could occur [11]. On one view, which we call value-dependent evidential reasoning , values should bias belief by affecting how people weigh evidence to form beliefs [11]. For instance, values could limit the kinds of evidence that justify certain beliefs (e.g., statistics versus testimony), or they could affect how strict or lenient one's evidential support must be to justify a belief (Box 2). On another view, moral and other values constitute independent grounds for belief and so act as value-based justi fi cations for belief [61,69,77]. For instance, a person might be justi fi ed in believing something they lack evidence to believe on the grounds that the belief is loyal (in the case of forming a favorable opinion about a friend) or useful (in the case of belief in God or free will).

These two proposals for how values might legitimately in fl uence belief formation can be roughly mapped onto two psychological models of biased reasoning. For instance, prominent models of belief formation posit that people accept (or reject) beliefs during inquiry once their evidence has passed a critical threshold [36]. On these models, thresholds for adopting (and rejecting) belief operate similarly to how they do for decision making (Box 1), in that they shift according to the costs of error, in this case, the costs associated with believing falsely. One way people bias this process is by holding desirable and undesirable beliefs to double-standards: submitting ' risky ' (undesirable) conclusions to more stringent evaluation and higher thresholds compared with ' safe ' (desirable) conclusions [16,36,78]. This form of biased reasoning may be justi fi ed according to some proposals of value-dependent evidential reasoning (Box 2).

## Box 2. Varieties of value-dependent evidential reasoning

There is an ongoing debate in epistemology regarding whether values ought to affect how people reason about evidence [11]. Even if people ought to be evidentialists (and so form beliefs only on suf fi cient evidence), values could in fl uence belief by affecting what counts as ' suf fi cient ' evidence. Here we review a few normative proposals in favor of value-dependent evidential reasoning and draw connections between these proposals and the lay ethics of belief.

One proposal is that the risks of making a wrong decision carry over to the risks of forming a false belief [76,96-99] (see also the concept of ' inductive risk ' in philosophy of science [100-102]). Accordingly, risks raise the ' evidential readiness level ' for decisions (Box 1) and for beliefs . For instance, if it would be bad for a scientist to publish an article wrongly suggesting that race correlates with IQ, then they ought to collect an extraordinary amount of evidence before doing so. Likewise, the scientist should acquire more evidence before believing (or ' accepting the hypothesis ' ) that race correlates with IQ. As noted in the main text, some work suggests that people believe evidence should be stricter for morally risky beliefs [10]. And related work suggests that, in risky situations, laypeople raise evidential thresholds for attributing knowledge to others [103,104] (but see [105]).

A closely related debate concerns whether people ought to form beliefs about others based on ' naked statistical evidence ' . For instance, ought someone judge that a well-dressed Black man in an exclusive club is a staff member merely because that is statistically likely [106]? Some argue ' no ' on the grounds that others have the right to avoid being wrongfully racially pro fi led [107] or the right to be treated as potential exceptions to generalizations [63-65]. Laypeople also often believe it is impermissible to judge others based on bare statistical information [10,108]. However, it is not clear whether people actually temper their impressions in this way [108-111].

The use of statistical evidence is also hotly discussed in the domain of judging others' guilt [112-114]. Some argue that, in order to judge someone guilty of some wrongdoing, people ought not rely solely on statistical evidence, but must obtain evidence that individualizes that person [115] (or has other properties [114,116]). Laypeople share these intuitions [117-119]; however, it is not clear why.

Emerging empirical work therefore suggests a variety of conditions under which people may endorse value-dependent evidential reasoning. However, questions remain about how often people do so, whether endorsing such reasoning affects belief formation, and fi nally, whether such reasoning is ultimately justi fi ed.

Other models of motivated reasoning posit that biased beliefs emerge in part because people value the social, moral, or practical value of a belief above the value of accuracy [27,79]. Accordingly, people adopt goals to acquire speci fi c beliefs and reason in ways that enable them to do so. This kind of motivated reasoning may be justi fi ed by theories according to which beliefs are valuable or not based on their moral and practical qualities, corresponding to value-based justi fi cation.

Studies on the lay ethics of belief, reviewed earlier, show that people sometimes endorse both value-dependent evidential reasoning and value-based justi fi cations for belief [10,32,33]. Corresponding to value-dependent evidential reasoning, people report that others need less evidence for morally bene fi cial beliefs, compared with morally risky ones, before judging that they have suf fi cient evidence [10]. For instance, in one study participants judged that a spouse had insuf fi cient evidence to form a pessimistic belief about his marriage but suf fi cient evidence to form an optimistic belief, even though they judged that, objectively, his evidence suggested a pessimistic outcome (i.e., divorce) (Figure 1). And, corresponding to value-based justi fi cations, the work reviewed earlier demonstrated that people often treat the positive effects of belief, such as whether a belief motivates someone, produces good behavior, or displays loyalty, as justi fi cations for belief even in the absence of suf fi cient evidence [10,32,33].

Taken together, these observations suggest that people may explicitly endorse reasoning that psychologists hypothesize gives rise to biased belief and this advocacy may be vindicated by normative theories of belief formation (Figure 2).

[Trends in Cognitive Sciences](Image of Figure 2)

Figure 2. Relationship between motivated reasoning, lay ethics of belief, and normative models of belief.

## Identifying promising interventions

Interventions that aim to replace biased reasoning with scienti fi c reasoning, what we call ' value debiasing ' interventions, face two challenges in light of the preceding discussion. First, if people reject the premise that they ought to be objective or that they ought to form beliefs on the basis of evidence, then they will reject interventions that merely instruct them to do so. And second, because many normative questions about how people ought to reason remain unsettled, interventions that promote scienti fi c reasoning may not result in ' better ' reasoning. Indeed, in some cases, people may not be biased enough , in which case interventions to be more objective would produce worse reasoning. For instance, it is possible that, for moral reasons, people ought to withhold forming beliefs about others based on race- or sex-based stereotypes, but that people rarely abide by these injunctions.

In light of these challenges, what sorts of interventions to induce scienti fi c reasoning might be defensible (normatively) and also effective (psychologically)? From the perspective of normative warrant, contexts in which it is clear that the potential objections to scienti fi c reasoning (featured in Table 1) do not apply are less likely to be problematic. From the perspective of psychological ef fi cacy, we propose that efforts to promote scienti fi c reasoning can leverage the commitments behind people's own ethic of belief. To this end, psychologists can ask which of the arguments for and against the scienti fi c ethos align with people's existing attitudes. Insofar as these attitudes lead people to devalue scienti fi c reasoning, we can subsequently ask: is this devaluation the result of a misconception? If so, interventions can target this misconception, with the expectation that scienti fi c reasoning will be valued more highly as a result [80,81]. This procedure offers a defensible and potentially effective way to improve reasoning despite uncertainty about what, ultimately, constitutes good reasoning in dif fi cult cases.

Asanexample, which we develop in Box 3, consider instrumental arguments for scienti fi c reasoning. According to such arguments, scienti fi c reasoning is valuable insofar as it maximizes welfare. This has the consequence that biased reasoning will be favored when it maximizes welfare instead. But if people are too liberal in their assumptions about the conditions under which biased reasoning

## Box 3. Case study: misconceptions about the practical bene fi ts of biased reasoning

Oneargumentfor the scienti fi c ethos is that scienti fi c reasoning is critical for optimal behavior and, therefore, for maximizing welfare (Table 1). In the context of value debiasing, this argument raises three questions. First, do people consider improving welfare to be a good foundation for norms? Yes . Second, do people devalue scienti fi c reasoning in part because they believe that biased reasoning will produce better outcomes? Yes [10,32]. And third, are people wrong to believe that biased reasoning will produce better outcomes? People may not be wrong in all cases [120,121]; for instance, some studies show that people bene fi t from holding biased beliefs about themselves [122] (but see [123]). Nevertheless, it is likely that people often sanction biased reasoning in error [84].

Even if inaccuracy accrues some minor advantage to believers, people tend to overestimate that advantage. In a recent series of studies, participants played short games and received either false information that they had performed well (and so should be optimistic about how well they will continue to perform) or that they had performed poorly (and so should be pessimistic) [32]. Across all tasks, including an age guessing game, a math test, and a Where's Waldo game, optimists never outperformed their pessimistic counterparts. However, a separate group of participants who were told about the experiment, exposed to the games and the manipulation, and fi nancially incentivized to make accurate predictions, expected the optimists to outperform the pessimists across all of these tasks. People were reliably unrealistic about the bene fi ts of unrealistic optimism.

People also appear to discount the bene fi ts of accuracy while overweighting the costs of honesty and transparency. For instance, people will provide inaccurate, overly positive feedback to others about their performance to avoid hurting their feelings [124], even though the long-term costs of inaccuracy likely outweigh the short-term negative feelings associated with criticism. In these situations, people may be overestimating the costs of hurtful, honest feedback [125]. If people were better calibrated to the costs of honesty, and to the long-term bene fi ts of fostering accurate beliefs in others, they would likely prioritize honesty more than they currently do.

Thus, educating people that unbiased belief and honesty align with human welfare more often than people currently anticipate appears to be a viable route to increasing their valuation of scienti fi c reasoning. After all, people already consider welfare a good reason to think unscienti fi cally; they just happen to hold demonstrably false beliefs about the effectiveness of bias in promoting welfare.

is productive, then, by their own lights, they are failing to implement their own values and would do better to engage in scienti fi c reasoning in a wider range of circumstances. Interventions might therefore focus on educating people about the norms for belief that in fact maximize welfare. Recent data also suggest other potentially promising routes to intervention. For instance, people likely often overestimate other people's ability to voluntarily change their beliefs [52]. This in turn may lead observers to demand biased beliefs from others when those others are unable to acquiesce (such as when a friend or spouse fails to adopt an overly favorable belief). Educating people about others' limited ability to adopt biased beliefs may reduce demand for such beliefs and, in so doing, reduce con fl ict that stems from people valuing beliefs based on their moral or other desirable qualities. And of course, future research may uncover new errors within people's lay theory of belief.

To our knowledge, no prior work has manipulated people's valuation of scienti fi c norms as a strategy for encouraging scienti fi c reasoning. Such interventions are promising for a few reasons. First, as noted earlier, there is a strong association between norms for belief and the kinds of beliefs people hold. Indeed, given that the association between the endorsement of scienti fi c norms and belief replicates across a wide range of beliefs, value debiasing has the potential to have similarly wide-ranging effects. Second, holding people accountable for their judgments affects how, and often how scienti fi cally, people reason [82]. In the present context, internalized norms of belief constitute a form of self-governance that may similarly affect an individual's own reasoning. At the same time, it is important to acknowledge some challenges. Many errors in reasoning are resistant to even large incentives to be accurate [83], suggesting that valuebased interventions may need to be accompanied by effective strategies for implementation. Thus, while value debiasing may re fl ect a promising starting point for improving reasoning, the most effective interventions will likely pair both information about why scienti fi c reasoning should be highly valued with tools and advice for implementing that reasoning [84].

## Trends in Cognitive Sciences

## Concluding remarks

It is no secret that humans are biased reasoners. Recent work suggests that these departures from scienti fi c reasoning are not simply the result of unconscious bias, but are also a consequence of endorsing norms for belief that place personal, moral, or social good above truth. The link between devaluing the ' scienti fi c ethos' and holding biased beliefs suggests that, in some cases, interventions on the perceived value of scienti fi c reasoning could lead to better reasoning and to better outcomes. In this spirit, we have offered a strategy for value debiasing. This strategy leverages work in epistemology to identify reasons for adopting or rejecting norms of belief and generating hypotheses for why people may devalue scienti fi c reasoning in some situations. By targeting and correcting errant assumptions underlying the devaluation of scienti fi c reasoning, psychologists may be able to increase the value people place on scienti fi c reasoning and therefore improve reasoning. There are two caveats to this approach (see Outstanding questions). First, biased reasoning is not necessarily bad reasoning. Resolving what constitutes justi fi ed or unjusti fi ed biased reasoning remains an important project for philosophers and psychologists. And second, boosting the perceived value of scienti fi c reasoning may not be suf fi cient to reason well. Nevertheless, a comprehensive approach to improving reasoning will bene fi t not only from educating people about what constitutes scienti fi c reasoning, but from af fi rming the value of scienti fi c reasoning in the fi rst place.

## Acknowledgments

We thank Andrew Chignell, Emily Liquin, S. Emlen Metz, Sarah McGrath, Ike Silver, Dan Singer, Michael Weisberg, and members of the Concepts and Cognition lab for helpful discussion and comments on previous drafts. We would also like to thank the Sense & Sensibility & Science team at UC Berkeley for some of the initial inspiration to pursue this work. The fi rst author is supported through funding provided by the Center for Human Values at Princeton University.

## Declaration of interests

No interests are declared.

## References

1. Baron, J. (2008) Thinking and Deciding , Cambridge University Press
2. Conee, E. and Feldman, R. (2004) Evidentialism , Oxford University Press
3. [Chisholm, R. (1957) Perceiving , Cornell University Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0015)
4. Shah, N. (2006) A new argument for evidentialism. Philos. Q. 56, 481-498
5. [Pinker, S. (2018) Enlightenment Now , Penguin Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0025)
6. [Merton, R.K. (1942) A note on science and democracy. J. Legal Polit. Sociol. 1, 115-126](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0030)
7. [Kunda, Z. (1990) The case for motivated reasoning. Psychol. Bull. 108, 480-498](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0035)
8. Kruglanski, A.W. (2004) The Psychology of Closed Mindedness , Psychology Press
9. [Pronin, E. et al. (2004) Objectivity in the eye of the beholder: divergent perceptions of bias in self versus others. Psychol. Rev. 111, 781-799](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0045)
10. [Cusimano, C. and Lombrozo, T. (2021) Morality justi fi es motivated reasoning in the folk ethics of belief. Cognition 209, 104513](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0050)
11. Bolinger, R.J. (2020) Varieties of moral encroachment. Philos. Perspect. Published online May 27, 2020. https://doi.org/ 10.1111/phpe.12124
12. [Koenig, M.A. et al. (2019) Children's judgments of epistemic and moral agents: from situations to intentions. Perspect. Psychol. Sci. 14, 334-360](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0060)
13. [Ståhl, T. et al. (2016) Moralized rationality: relying on logic and evidence in the formation and evaluation of belief can be seen as a moral issue. PLoS One 11, e0166332](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0065)
14. [Ross, L. and Ward, A. (1996) Naive realism in everyday life: Implications for social con fl ict and misunderstanding. In The Jean Piaget Symposium Series. Values and Knowledge (Reed, E.S. et al ., eds), pp. 103-135, Lawrence Erlbaum Associates](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0070)
15. [Nickerson, R.S. (1998) Con fi rmation bias: a ubiquitous phenomenon in many guises. Rev. Gen. Psychol. 2, 175-220](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0075)
16. [Ditto, P.H. and Lopez, D.F. (1992) Motivated skepticism: use of differential decision criteria for preferred and nonpreferred conclusions. J. Pers. Soc. Psychol. 63, 568-584](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0080)
17. [Pennycook, G. et al. (2015) On the reception and detection of pseudo-profound bullshit. Judgm. Decis. Mak. 10, 549-563](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0085)
18. [Kahneman, D. and Tversky, A. (1972) Subjective probability: a judgment of representativeness. Cogn. Psychol. 3, 430-454](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0090)
19. [Ehrlinger, J. et al. (2005) Peering into the bias blind spot: people's assessments of bias in themselves and others. Personal. Soc. Psychol. Bull. 31, 680-692](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0095)
20. [Fischhoff, B. (1982) Debiasing. In Judgment under Uncertainty: Heuristics and Biases (Kahneman, D. et al ., eds), pp. 422-444, Cambridge University Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0100)
21. [Sellier, A.L. et al. (2019) Debiasing training improves decision making in the fi eld. Psychol. Sci. 30, 1371-1379](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0105)
22. [Fong, G.T. et al. (1986) The effects of statistical training on thinking about everyday problems. Cogn. Psychol. 18, 253-292](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0110)
23. [Fong, G.T. and Nisbett, R.E. (1991) Immediate and delayed transfer of training effects in statistical reasoning. J. Exp. Psychol. Gen. 120, 34-45](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0115)
24. [Milkman, K.L. et al. (2009) How can decision making be improved. Perspect. Psychol. Sci. 4, 379-383](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0120)
25. [Barnett, S.M. and Ceci, S.J. (2002) When and where do we apply what we learn? A taxonomy for far transfer. Psychol. Bull. 128, 612-637](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0125)
26. [Willingham, D.T. (2008) Critical thinking: why is it so hard to teach. Arts Educ. Policy Rev. 109, 21-32](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0130)
27. [Tetlock, P.E. (2002) Social functionalist frameworks for judgment and choice: intuitive politicians, theologians, and prosecutors. Psychol. Rev. 109, 451-471](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0135)

## Outstanding questions

When, if ever, does biased reasoning improve welfare? A large literature has tackled this question in the domain of self-assessment, but recent results suggest an answer is far off.

Do any of the four common arguments for scienti fi c reasoning (Table 1) offer the basis for effective interventions on lay attitudes?

Is manipulating the perceived value of scienti fi c reasoning suf fi cient to improve reasoning? If not, what else is required?

Can psychologists develop ' value debiasing ' interventions that generalize across situations? Cognitive training interventions have had dif fi culty generalizing outside the training context, but interventions that target the perceived value of scienti fi c reasoning may have the potential to generalize more broadly.

Some work documents individual differences in valuation of scienti fi c reasoning. Absent intervention, what gives rise to low or high valuation of scienti fi c reasoning?

Are people who score highly in dispositional acceptance and valuation of scienti fi c reasoning less likely to accept moral and social arguments in favor of biased reasoning? Or do (some) moral or social cases constitute a special exception?

What are limits on people's morally motivated reasoning? Are people more successful at acquiring motivated beliefs for moral reasons compared with other reasons, or does unambiguous evidence constrain morally motivated reasoning, too?

We have focused on a subset of the norms considered central to the scienti fi c ethos, namely those concerned with evidence and impartiality. What role, if any, do other norms, such as universalism and communalism, play in the lay ethics of belief?

Are there cases in which people are less biased than they ought to be? Answering this question requires settling both normative and empirical matters.

28. [Cohen, L.J. (1981) Can human irrationality be experimentally demonstrated. Behav. Brain Sci. 4, 317-331](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0140)
29. [Keys, D.J. and Schwartz, B. (2007) ' Leaky ' rationality: how research on behavioral decision making challenges normative standards of rationality. Perspect. Psychol. Sci. 2, 162-180](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0145)
30. [Armor, D.A. et al. (2008) Prescribed optimism: is it right to be wrong about the future. Psychol. Sci. 19, 329-331](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0150)
31. [Miller, J.E. et al. (2021) Do people prescribe (over)optimism. Psychol. Sci. (in press)](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0155)
32. [Tenney, E.R. et al. (2015) (Too) optimistic about optimism: the belief that optimism improves performance. J. Pers. Soc. Psychol. 108, 377-399](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0160)
33. [Metz, S.E. et al. (2018) Non-scienti fi c criteria for belief sustain counter-scienti fi c beliefs. Cogn. Sci. 42, 1477-1503](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0165)
34. [Epley, N. and Gilovich, T. (2016) The mechanics of motivated reasoning. J. Econ. Perspect. 30, 133-140](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0170)
35. [von Hippel, W. and Trivers, R. (2011) The evolution and psychology of self-deception. Behav. Brain Sci. 34, 1-16](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0175)
36. [Trope, Y. and Liberman, A. (1996) Social hypothesis-testing: cognitive and motivational mechanisms. In Social Psychology: Handbook of Basic Principles (Higgins, E.T. and Kruglanski, A.W., eds), pp. 239-270, The Guilford Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0180)
37. [Adam-Troian, J. et al. (2019) Unfounded beliefs among teachers: the interactive role of rationality priming and cognitive ability. Appl. Cogn. Psychol. 33, 720-727](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0185)
38. [Pennycook, G. et al. (2020) On the belief that beliefs should change according to evidence: implications for conspiratorial, moral, paranormal, political, religious, and science beliefs. Judgm. Decis. Mak. 15, 476-498](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0190)
39. [Ståhl, T. and van Prooijen, J.-W. (2018) Epistemic rationality: skepticism toward unfounded beliefs requires suf fi cient cognitive ability and motivation to be rational. Personal. Individ. Differ. 122, 155-163](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0195)
40. [Ståhl, T. and van Prooijen, J.-W. (2021) Analytic atheism: valuing epistemic rationality strengthens the association between analytic thinking and religious disbelief. Personal. Individ. Differ. 179, 110914](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0200)
41. [Stenhouse, N. et al. (2018) The potential role of actively openminded thinking in preventing motivated reasoning about controversial science. J. Environ. Psychol. 57, 17-24](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0205)
42. [Stanovich, K.E. and Toplak, M.E. (2019) The need for intellectual diversity in psychological science: our own studies of actively open-minded thinking as a case study. Cognition 187, 156-166](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0210)
43. [Stanovich, K.E. and West, R.F. (1997) Reasoning independently of prior belief and individual differences in actively open-minded thinking. J. Educ. Psychol. 89, 342-357](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0215)
44. [Stanovich, K.E. and West, R.F. (1998) Individual differences in rational thought. J. Exp. Psychol. Gen. 127, 161-188](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0220)
45. [Toplak, M.E. et al. (2017) Real-world correlates of performance on heuristics and biases tasks in a community sample. J. Behav. Decis. Mak. 30, 541-554](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0225)
46. [Haran, U. et al. (2013) The role of actively open-minded thinking in information acquisition, accuracy, and calibration. Judgm. Decis. Mak. 8, 188-201](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0230)
47. [Mellers, B. et al. (2015) The psychology of intelligence analysis: drivers of prediction accuracy in world politics. J. Exp. Psychol. Appl. 21, 1-14](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0235)
48. [Baron, J. (2019) Actively open-minded thinking in politics. Cognition 188, 8-18](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0240)
49. [Carpenter, J. et al. (2018) The impact of actively open-minded thinking on social media communication. Judgm. Decis. Mak. 13, 562-574](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0245)
50. [Bronstein, M.V. et al. (2019) Belief in fake news is associated with delusionality, dogmatism, religious fundamentalism, and reduced analytic thinking. J. Appl. Res. Mem. Cogn. 8, 108-117](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0250)
51. [Cusimano, C. and Goodwin, G.P. (2019) Lay beliefs about the controllability of everyday mental states. J. Exp. Psychol. Gen. 148, 1701-1732](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0255)
52. [Cusimano, C. and Goodwin, G.P. (2020) People judge others to have more voluntary control over beliefs than they themselves do. J. Pers. Soc. Psychol. 119, 999-1029](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0260)
53. [Tetlock, P.E. et al. (1989) Social and cognitive strategies for coping with accountability: conformity, complexity, and bolstering. J. Pers. Soc. Psychol. 57, 632-640](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0265)
54. Williams, D. (2020) Socially adaptive belief. Mind Lang. 36, 333-354
55. [Anderson, C. et al. (2012) A status-enhancement account of overcon fi dence. J. Pers. Soc. Psychol. 103, 718-735](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0275)
56. [Henrich, J. and Gil-White, F.J. (2001) The evolution of prestige: freely conferred deference as a mechanism for enhancing the bene fi ts of cultural transmission. Evol. Hum. Behav. 22, 165-196](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0280)
57. [Cheng, J.T. et al. (2021) The social transmission of overcon fi dence. J. Exp. Psychol. Gen. 150, 157-186](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0285)
58. Sulik, J. et al. (2021) Collectively jumping to conclusions: social information ampli fi es the tendency to gather insuf fi cient data. J. Exp. Psychol. Gen. Published online March 18, 2021. http://doi.org/10.1037/xge0001044
59. [De Dreu, C.K. et al. (2008) Motivated information processing in group judgment and decision making. Personal. Soc. Psychol. Rev. 12, 22-49](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0295)
60. [Nelson, L.D. et al. (2018) Psychology's renaissance. Annu. Rev. Psychol. 69, 511-534](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0300)
61. Stroud, S. (2006) Epistemic partiality in friendship. Ethics 116, 498-524
62. [Gagné, F.M. and Lydon, J.E. (2004) Bias and accuracy in close relationships: an integrative review. Personal. Soc. Psychol. Rev. 8, 322-338](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0310)
63. [Armour, J.D. (1994) Race ipsa loquitur: of reasonable racists, intelligent Bayesians, and involuntary negrophobes. Stanford Law Rev. 46, 781](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0315)
64. [Basu, R. (2019) What we epistemically owe to each other. Philos. Stud. 176, 915-931](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0320)
65. Moss, S. (2018) Probabilistic Knowledge , Oxford University Press
66. [Atran, S. and Norenzayan, A. (2004) Religion's evolutionary landscape: counterintuition, commitment, compassion, communion. Behav. Brain Sci. 27, 713-730](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0330)
67. [Boyer, P. (2001) Religion Explained: The Human Instincts That Fashion Gods , Basic Books, Spirits and Ancestors](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0335)
68. [James, W. (1937) The Will to Believe, and Other Essays in Popular Philosophy , Longmans, Green and Co](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0340)
69. [Pascal, B. (1852) Pensées , Dezobry et E. Magdeleine](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0345)
70. [Clifford, W.K. (1877) The ethics of belief. In The Ethics of Belief and Other Essays (Madigan, T., ed.), pp. 70-96, Prometheus](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0350)
71. Foley, R. (1987) The Theory of Epistemic Rationality , Harvard University Press
72. [Kornblith, H. (1993) Epistemic normativity. Synthese 94, 357-376](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0360)
73. [Chisholm, R.M. (1989) Theory of Knowledge , Prentice Hall](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0365)
74. [Zagzebski, L.T. (1996) Virtues of the Mind: An Inquiry Into the Nature of Virtue and the Ethical Foundations of Knowledge , Cambridge University Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0370)
75. [Côté-Bouchard, C. (2017) Is epistemic normativity valuebased. Dialogue 56, 407-430](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0375)
76. [Pace, M. (2011) The epistemic value of moral considerations: justi fi cation, moral encroachment, and James ' ' will to believe ' . Noûs 45, 239-268](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0380)
77. Basu, R. (2019) The wrongs of racist beliefs. Philos. Stud. 176, 2497-2515
78. [Gilovich, T. (1991) How We Know What Isn't So: The Fallibility of Human Reason in Everyday Life , Free Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0390)
79. [Van Bavel, J.J. and Pereira, A. (2018) The partisan brain: an identity-based model of political belief. Trends Cogn. Sci. 22, 213-224](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0395)
80. [Rokeach, M. (1971) Long-range experimental modi fi cation of values, attitudes, and behavior. Am. Psychol. 26, 453-459](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0400)
81. [Walton, G.M. and Wilson, T.D. (2018) Wise interventions: psychological remedies for social and personal problems. Psychol. Rev. 125, 617-655](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0405)
82. [Lerner, J.S. and Tetlock, P.E. (1999) Accounting for the effects of accountability. Psychol. Bull. 125, 255-275](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0410)
83. [Enke, B. et al. (2021) Cognitive Biases: Mistakes or Missing Stakes? (28650) , National Bureau of Economic Research](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0415)
84. [Moore, D.A. (2020) Perfectly Con fi dent , Harper Business](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0420)
85. [Douglas, H. (2016) Values in science. In Oxford Handbook of Philosophy of Science Values in Science (Humphreys, P., ed.), pp. 609-630, Oxford University Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0425)
86. [Laudan, L. (2004) The epistemic, the cognitive, and the social. In Science, Values, and Objectivity (Machamer, P. and Wolters, G., eds), pp. 14-23, University of Pittsburgh Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0430)

## Trends in Cognitive Sciences

87. [IJzerman, H. et al. (2020) Use caution when applying behavioural science to policy. Nat. Hum. Behav. 4, 1092-1094](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0435)
88. [Harackiewicz, J.M. et al. (2016) Closing achievement gaps with a utility-value intervention: disentangling race and social class. J. Pers. Soc. Psychol. 111, 745-765](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0440)
89. [Hulleman, C.S. and Harackiewicz, J.M. (2009) Promoting interest and performance in high school science classes. Science 326, 1410-1412](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0445)
90. [Yeager, D.S. et al. (2014) Boring but important: a selftranscendent purpose for l earning fosters academic self-regulation. J. Pers. Soc. Psychol. 107, 559-580](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0450)
91. [Dubey, R. and Grif fi ths, T.L. (2020) Reconciling novelty and complexity through a rational analysis of curiosity. Psychol. Rev. 127, 455-476](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0455)
92. [Dubey, R. et al. (2019) If it's important, then I am curious: a value intervention to induce curiosity. In Proceedings of the 41st Annual Conference of the Cognitive Science Society , pp. 282-288](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0460)
93. [Liquin, E.G. and Lombrozo, T. (2020) A functional approach to explanation-seeking curiosity. Cogn. Psychol. 119, 101276](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0465)
94. [Mayseless, O. and Kruglanski, A.W. (1987) What makes you so sure? Effects of epistemic motivations on judgmental con fi dence. Organ. Behav. Hum. Decis. Process. 39, 162-183](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0470)
95. [McAllister, D.W. et al. (1979) The contingency model for the selection of decision strategies: an empirical test of the effects of signi fi cance, accountability, and reversibility. Org. Behav. Hum. Perform. 24, 228-244](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0475)
96. [Fantl, J. and McGrath, M. (2002) Evidence, pragmatics, and justi fi cation. Philos. Rev. 111, 67-94](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0480)
97. [Weatherson, B. (2005) Can we do without pragmatic encroachment. Philos. Perspect. 19, 417-443](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0485)
98. [Ross, J. and Schroeder, M. (2014) Belief, credence, and pragmatic encroachment. Philos. Phenomenol. Res. 88, 259-288](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0490)
99. [Basu, R. (2019) Radical moral encroachment: the moral stakes of racist beliefs. Philos. Issues 29, 9-23](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0495)
100. Douglas, H. (2000) Inductive risk and values in science. Philos. Sci. 67, 559-579
101. [Elliott, K.C. (2011) Direct and indirect roles for values in science. Philos. Sci. 78, 303-324](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0505)
102. [Rudner, R. (1953) The scientist qua scientist makes value judgments. Philos. Sci. 20, 1-6](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0510)
103. Dinges, A. and Zakkou, J. (2020) Much at stake in knowledge. Mind Lang. Published online June 8, 2020. https://doi.org/ 10.1111/mila.12300
104. [Pinillos, Á. (2012) Knowledge, experiments, and practical interests. In Knowledge Ascriptions (Brown, J. and Gerken, M., eds), pp. 192-219, Oxford University Press](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0520)
105. Rose, D. et al. (2019) Nothing at stake in knowledge. Noûs 53, 224-247
106. [Gendler, T.S. (2011) On the epistemic costs of implicit bias. Philos. Stud. 156, 33-63](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0530)
107. [Bolinger, R.J. (2020) The rational impermissibility of accepting (some) racial generalizations. Synthese 197, 2415-2431](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0535)
108. [Cao, J. et al. (2019) People make the same Bayesian judgment they criticize in others. Psychol. Sci. 30, 20-31](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0540)
109. [Cao, J. and Banaji, M.R. (2016) The base rate principle and the fairness principle in social judgment. Proc. Natl. Acad. Sci. U. S. A. 113, 7475-7480](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0545)
110. [Yzerbyt, V.Y. et al. (1994) Social judgeability: the impact of meta-informational cues on the use of stereotypes. J. Pers. Soc. Psychol. 66, 48-55](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0550)
111. [Jost, J.T. et al. (1998) Social metacognition: an expansionist review. Personal. Soc. Psychol. Rev. 2, 137-154](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0555)
112. [Buchak, L. (2014) Belief, credence, and norms. Philos. Stud. Int. J. Philos. Anal. Tradit. 169, 285-311](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0560)
113. [Tribe, L.H. (1971) Trial by mathematics: precision and ritual in the legal process. Harvard Law Rev. 84, 1329-1393](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0565)
114. [Smith, M. (2018) When does evidence suf fi ce for conviction. Mind 127, 1193-1218](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0570)
115. [Bolinger, R.J. (2021) Explaining the justi fi catory asymmetry between statistical and individualized evidence. In The Social Epistemology of Legal Trials (Hoskins, Z. and Robson, J., eds), pp. 60-76, Routledge](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0575)
116. [Enoch, D. et al. (2012) Statistical evidence, sensitivity, and the legal value of knowledge. Philos Public Aff. 40, 197-224](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0580)
117. [Arkes, H.R. et al. (2012) Disjunction between probability and verdict in juror decision making. J. Behav. Decis. Mak. 25, 276-294](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0585)
118. [Niedermeier, K.E. et al. (1999) Jurors' use of naked statistical evidence: exploring bases and implications of the wells effect. J. Pers. Soc. Psychol. 76, 533-542](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0590)
119. [Wells, G.L. (1992) Naked statistical evidence of liability: is subjective probability enough. J. Pers. Soc. Psychol. 62, 739-752](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0595)
120. [Baumeister, R.F. (1989) The optimal margin of illusion. J. Soc. Clin. Psychol. 8, 176-189](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0600)
121. [Carver, C.S. and Scheier, M.F. (2014) Dispositional optimism. Trends Cogn. Sci. 18, 293-299](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0605)
122. [He, J.C. and Côté, S. (2019) Self-insight into emotional and cognitive abilities is not related to higher adjustment. Nat. Hum. Behav. 3, 867-884](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0610)
123. [Humberg, S. et al. (2019) Is accurate, positive, or in fl ated selfperception most advantageous for psychological adjustment? A competitive test of key hypotheses. J. Pers. Soc. Psychol. 116, 835-859](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0615)
124. [Lupoli, M.J. et al. (2017) Lying because we care: compassion increases prosocial lying. J. Exp. Psychol. Gen. 146, 1026-1042](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0620)
125. [Levine, E.E. and Cohen, T.R. (2018) You can handle the truth: mispredicting the consequences of honest communication. J. Exp. Psychol. Gen. 147, 1400-1429](http://refhub.elsevier.com/S1364-6613(21)00151-0/rf0625)

## Figures (extracted)

*Figure images auto-extracted from the PDF via Docling; captions as published.*

**Figure I. The relationship between value, inquiry, and belief in the scienti fi c ethos.**
![Figure I. The relationship between value, inquiry, and belief in the s](figures/cusimano_lombrozo_2021_tics/fig01.png)

*Figure description (AI-generated from the image; qualitative, not a substitute for the reported statistics):* A conceptual flow diagram (not data) of how social and moral values can legitimately shape inquiry without directly dictating belief. Values feed three channels of "value-guided inquiry": value-based question-asking ("What is valuable to know?", e.g. broader-impacts statements), costs of knowledge ("Is acquiring knowledge worth the trade-offs?", e.g. Institutional Review Boards), and due diligence ("How much certainty is required for action?", e.g. readiness levels). These shape the evidence that inquiry yields, which in turn feeds belief formation -- while a "scientific ethos" blocks values from affecting belief directly.