This study presents findings from a 4-year panel study examining three major questions regarding the measurement of social value orientation (SVO). First, we investigate the test–retest reliability of the Slider Measure (SLM, Murphy et al., 2014) over a period of up to 4 years in a large, demographically diverse sample. Second, we compare the stability of the SLM to related measurements of prosociality and distributional preferences along the behavior–behavioral tendency–trait continuum, including single behaviors (e.g., the Dictator Game and the Prisoner’s Dilemma), alternative behavioral tendencies (e.g., survey-based measures of altruism), and broader personality traits (e.g., Big-Five, HEXACO, Dark Factor D). Third, we explore differences in individual trajectories of SVO, focusing on how age and gender influence its stability and change over time. Our study thus complements earlier research on the stability of the SLM by extending the time period and depth of analysis, and putting the measure in the context of other related measures. The results show a considerable degree of stability, higher than all behavioral games, but often lower than fully fledged measures of personality traits. Furthermore, we find that age has a stabilizing effect on behavior in the SLM. With regard to gender, we find that women behave generally more prosocial than men but that they do not differ in their stability. We conclude that the SLM is a suitable method for assessing individual SVO over longer time periods and is best thought of as covering a sweet spot between stable personality traits and immediate behavioral expressions.

People who repeatedly risk money, whether they be traders for financial institutions, corporate executives, day traders, or sports bettors, sometimes appear to do better than chance only because the risk of large losses is hidden or overlooked. As students of casino gambling know, one way to obscure the risk of large losses is to bet more when you are losing and less when you are winning. In 19th century casinos, betting strategies that did this were called martingales. Following such strategies, whether deliberately or unwittingly, was called martingaling. Traders in financial instruments often martingale; in fact, they are martingaling whenever they respond to a margin call. A businessperson who doubles down on an apparently losing investment is martingaling. Opinionated sports bettors easily fall into martingaling. The martingale index, defined in this paper, measures the portion of the apparent success of a betting, trading, or investment strategy that can be attributed to martingaling. We calculate the martingale index for some popular casino strategies and also for some strategies that model random trading in S&P 500 futures and in stocks. And we discuss how educating the public about the martingale index might help both businesses and individuals avoid the temptations of martingaling.

Some scholars have treated overconfidence as an individual difference—that is, assuming the tendency to be overconfident is stable within a person and differs meaningfully from person to person. We question this assumption. We investigate consistency within individuals between its three forms—overestimation, overplacement, and overprecision—in multiple domains (Study 1a and 1b), at multiple times (Study 1b and 2), and with multiple measures (Study 3a and 3b). We find mixed evidence of trait-like consistency. We do find some evidence of within-individual stability across domains and time points. However, we find little consistency across different measures of the same form of overconfidence—specifically overprecision. Instead, we find more consistent evidence that overconfidence varies situationally and contextually.

In four studies, participants judged satisfaction with hypothetical salaries, given the salaries of others doing the same work. Unlike previous research, contexts (distributions of others’ salaries) were manipulated within- rather than between-subjects. These studies enabled tests of an extension of range–frequency (RF) theory that assumes that judgments are a compromise between RF predictions based on between- and within-trial contexts. This extension to within-subjects designs correctly predicted the cases in which people assign higher satisfaction ratings to lower salaries. The manipulation of the context within-subjects confirmed phenomena previously observed in between-subjects research. However, a violation of this within-subjects RF model was also observed: When one’s salary is lowest compared to others within the same firm, satisfaction varies inversely with the highest salary paid to another at the same firm. Apparently, judgments of satisfaction also depend on inequity. This finding was not observed in previous between-subjects research; indeed, salary and inequity are perfectly confounded for the participant in such a design. We theorize that satisfaction is not merely a judgment of where one’s salary falls relative to other salaries, but also depends on how much one is underpaid relative to the distribution of underpayments. A revision of the within-subjects RF model (incorporating the distribution of inequities) gave a good description of judgments of salary satisfaction and of the likelihood to accept a job offer.

Large language models (LLMs) such as ChatGPT, Gemini, and Claude are increasingly being used in aid or place of human judgment and decision making. Indeed, academic researchers are increasingly using LLMs as a research tool. In this paper, we examine whether LLMs, like academic researchers, fall prey to a particularly common human error in interpreting statistical results, namely ‘dichotomania’ that results from the dichotomization of statistical results into the categories ‘statistically significant’ and ‘statistically nonsignificant’. We find that ChatGPT, Gemini, and Claude fall prey to dichotomania at the 0.05 and 0.10 thresholds commonly used to declare ‘statistical significance’. In addition, prompt engineering with principles taken from an American Statistical Association Statement on Statistical Significance and P-values intended as a corrective to human errors does not mitigate this and arguably exacerbates it. Further, more recent and larger versions of these models do not necessarily perform better. Finally, these models sometimes provide interpretations that are not only incorrect but also highly erratic.

Prior research has shown that people judge algorithmic errors more harshly than identical mistakes made by humans—a bias known as algorithm aversion. We explored this phenomenon across two studies (N = 1199), focusing on the often-overlooked role of conventionality when comparing human versus algorithmic errors by introducing a simple conventionality intervention. Our findings revealed significant algorithm aversion when participants were informed that the decisions described in the experimental scenarios were conventionally made by humans. However, when participants were told that the same decisions were conventionally made by algorithms, the bias was significantly reduced—or even completely offset. This intervention had a particularly strong influence on participants’ recommendations of which decision-maker should be used in the future—even revealing a bias against human error makers when algorithms were framed as the conventional choice. These results suggest that the existing status quo plays an important role in shaping people’s judgments of mistakes in human–algorithm comparisons.

Graphs can help people arrive at data-supported conclusions. However, graphs might also induce bias by shifting the amount of evidence needed to make a decision, such as deciding whether a treatment had some kind of effect. In 2 experiments, we manipulated the early base rates of treatment effects in graphs. Early base rates had a large effect on a signal detection measure of bias in future graphs even though all future graphs had a 50% chance of showing a treatment effect, regardless of earlier base rates. In contrast, the autocorrelation of data points within each graph had a larger effect on discriminability. Exploratory analyses showed that a simple cue could be used to correctly categorize most graphs, and we examine participants’ use of this cue among others in lens models. When exposed to multiple graphs on the same topic, human judges can draw conclusions about the data, but once those conclusions are made, they can affect subsequent graph judgment.

Past research has documented ingroup favoritism, the tendency to cooperate more with ingroup members than outgroup members, in a wide range of intergroup contexts, and extensively discussed conditions under which ingroup favoritism emerges. However, previous studies have predominantly focused on a simplistic intergroup context, for instance, where group boundaries are static, and one group membership is present. To fill the gap, we leveraged data from professional volleyball players and investigated the influence of (1) varying levels of intergroup conflict salience, (2) past and present group memberships, and (3) national team membership on intergroup cooperation. Contrary to our hypotheses and the social identity perspective, we found that conflict salience and former ingroup membership did not influence intergroup cooperation. Additionally, we found that the more national team players there are in the ingroup, the more cooperative those who play for the national team are with ingroup members, leading to increased ingroup favoritism.

Trust is essential for effective collaboration. In advice settings, decision-makers’ trust in their advisors determines their willingness to follow advice. We propose that trust in the opposite direction, that is, the trust of the advisor in the decision-maker, can affect the use of advice. Specifically, we suggest that advice-taking is greater after a show of trust by the advisor than after an instance of distrust. We conducted four behavioral experiments using the trust game and judge–advisor system paradigms and one scenario study using a sample of currently employed professionals (N = 1599). We find that initial displays of trust by advisors result in greater acceptance of their advice (Studies 1A-B). This effect persists across different levels of advice quality, resulting in smaller underutilization of high-quality advice but also in overreliance on low-quality advice (Study 2). Decision-makers not only show greater willingness to follow advisors who trust them but also respond similarly to advisors who display trust in other people (Study 3). Finally, we find evidence for both perceived advisor competence and decision-makers’ motivation to reciprocate as mediators of the relation between advisors’ level of trust and decision-makers’ willingness to follow their advice (Study 4). Our findings shed light on the dynamics of trust and persuasion in advice relationships and provide insight for advisors who wish to maintain the effectiveness of their input.

Risk aversion—the preference for certainty over potential gains or losses—is reduced when using a foreign language. We investigated 2 mechanisms for this foreign language effect using incentivized gambles with verbal probability expressions: (1) that people perceive prospects of winning as larger when a decision is made in their foreign language; (2) that people experience reduced negative affect toward risk in a foreign language. In Experiment 1, N = 229 proficient Polish–English bilinguals, using ungridded slider, assigned numerical values to 29 verbal probability expressions in both languages. We found small bidirectional differences in 13 of them, leading us to reject the first mechanism. In Experiment 2, N = 281 participants gambled in incentivized neutral expected value lotteries using a sample of the verbal probability expressions from Experiment 1. Participants gambled in either their native or foreign language, where participants could either accept around 50% of gambles (debiased to risk-neutral) or more than 50% (biased to risk-seeking). Surprisingly, we observed no significant risk aversion in either language condition, with participants’ gambling behavior close to 50% in both cases. The finding that participants showed no risk aversion in native language condition meant we could not test whether foreign language reduces risk aversion. However, this result did show that using a foreign language does not promote excessive risk-taking. Our findings suggest that using verbal probability expressions does not bias participants’ responses, and may inherently reduce risk aversion.

There are large individual differences in prosocial vs. antisocial behavior as studied via economic games. Prominent among the personality traits that have been considered as potential correlates are ‘dark’ traits (especially Machiavellianism, Narcissism, Psychopathy, and Sadism). However, although such traits should account for choices in games, the corresponding associations are weak and inconsistent, leading to a state of knowledge that lacks specificity and parsimony. We argue and demonstrate across 10 studies and 8 economic games (total N = 10,474) that a clearer picture emerges once (a) including games that (also) afford the expression of forgiveness (vs. retaliation) and/or (dis)trust and (b) considering the shared vs. unique aspects of dark traits. Specifically, we find that (i) the common core of all dark traits—the dark factor of personality, D—consistently predicts antisocial behavior in all games, (ii) dark traits also (though less strongly) predict antisocial behavior, and (iii) dark traits do so almost entirely due to D. We discuss the theoretical and methodological implications of these findings for the study of individual differences in pro- vs. antisocial behavior in economic games.

To the known causes of overconfidence in decisions and judgments, we reveal another source that derives from a bias during the act of decision making. While this bias, the predecisional distortion of information, is well studied, its impact on overconfidence is not. We demonstrate how the distortion of information creates overconfidence in those professionals often regarded as singularly overconfident, entrepreneurs. When these professionals use a sequence of relevant information to make an accept-reject decision about a business opportunity, a cycle of confidence-distortion-confidence builds unjustified confidence in the chosen action – and does so whether that action is to accept or reject the venture. Overconfidence is a well-recognized cause of flawed decision making. Our work demonstrates the paradoxical converse of this claim, that flawed decision making can be a cause of overconfidence.

In 2 experiments, we attempted to reduce belief-consistent biases in interpretations of a polarized problem by making information easier to interpret. In the experiments, participants solved numerical problems that were either framed in a politically polarized (the effects of Muslim prayer rooms on support for Islamic extremism) or a neutral setting (the effects of a skin cream on skin rash). In both studies, the problems were presented twice, with the second presentation accompanied with an aid to facilitate problem-solving. In Experiment 1, this aid came in the form of an informative text on how to calculate the numbers to solve the problem. In Experiment 2, the aid provided participants with the first calculus necessary to solve the problem: transforming frequencies to percentages. Overall, results demonstrated belief-consistent responses in the polarized scenario when participants attempted to solve the first problem (higher accuracy when the correct conclusion was in line with participants’ ideology). Information on how to calculate the problem (Experiment 1) only slightly reduced the biased responses, whereas the added percentages (Experiment 2) led to a substantial reduction of the bias. Thus, we demonstrate that the facilitation of complex information on a polarized topic reduces biases in favor of rational reasoning.

Existing research suggests a negative correlation between reflective thinking and religious belief. The dual process model (DPM) posits that reflection diminishes religious belief by limiting intuitive decisions. In contrast, the expressive rationality model (ERM) argues that reflection serves an identity-protective function by bolstering rather than modifying preexisting beliefs. Although the current literature tends to favor the DPM, many studies suffer from unbalanced samples. To avoid this limitation, we recruited comparably large number of participants for both religious believers (n = 580) and non-believers (n = 594) and observed the relationship between reflection and two measures of religious belief: belief in God and disbelief in evolution. Our findings corroborate the negative associations found between higher levels of reflection and both types of belief, independent of religious affiliation. Our results align with the broader literature, supporting the DPM but not the ERM.

Addressing climate change requires substantial shifts in individual behavior. Providing information about climate externalities through carbon labels is a promising tool to foster sustainable choices when individuals weigh environmental against personal outcomes. We study the impact of carbon labels over time and their underlying mechanisms in a repeated risky choice task. We ran two preregistered experiments (US samples, N = 1,268) with realized carbon and monetary payoffs, examining how choice is influenced by the timing of carbon information (One-off vs. Recurring) and participants’ political preference (i.e., Democrat or Republican voters). In Study 1, we find that both Democrats and Republicans reduce carbon emissions when carbon labels were provided. Further, recurring labels significantly reduced carbon choices compared to one-off labels. Study 2 replicated the results in a within-participant design and showed that the impact of recurring carbon labels on sustainable choices cannot be explained by the strength of emission recall. This suggests that recurring labels amplify the importance given to the climate attribute in the decision process, operating via attentional rather than informational mechanisms. Our results emphasize the importance of providing climate externality information at time of use to raise awareness about climate costs and bolster sustainable preferences across population segments.

There is a large literature evaluating the dual process model of cognition, including the biases and heuristics it implies. However, our understanding of what causes effortful thinking remains incomplete. To advance this literature, we focus on what triggers decision-makers to switch from the intuitive process (System 1) to the more deliberative process (System 2). We examine how the framing of incentives (gains versus losses) influences decision processing. To evaluate this, we design experiments based on a task developed to distinguish between intuitive and deliberative thinking. Replicating previous research, we find that losses elicit more cognitive effort. Most importantly, we also find that losses differentially reduce the incidence of intuitive answers, consistent with triggering a shift between these modes of cognition. We find substantial heterogeneity in these effects, with young men being much more responsive to the loss framing. To complement these findings, we provide robustness tests of our results using aggregated data, the imposition of a constraint to hinder the activation of System 2, and an analysis of incorrect, but unintuitive, answers to inform hybrid models of choice.

This research explores a phenomenon that we see nearly every day and has implications for how we view people in other nations: Different media outlets may report the same international events either in terms of the nation (e.g., “Russia invades Ukraine”) or in terms of the leader (e.g., “Putin invades Ukraine”). Five studies, conducted during the 2022 Russia-Ukraine Conflict and involving both field and experimental data, find that readers of nation-framed news about the conflict had worse impressions of the people in the associated nation (Russians) than readers of the corresponding leader-framed version. We explain the psychology behind this framing effect and identify its moderators. Our research underscores the importance of responsible media practices in shaping global perceptions.

Many people struggle to make tradeoffs between present wants and future wishes, resulting in the tendency to overly discount the value of future rewards. To explain such behavior, past work has pointed to future self-continuity or the perceived connection between a person’s current and future selves. Yet, most of this past work has been conducted on small-to-medium size convenience samples, and as such, little is known regarding the population-level statistics of future self-continuity or how its link to important financial health variables like saving behavior and financial well-being play out in a nationally representative sample. Here, we use a nationally representative sample of over 6,000 Americans to investigate the generalizability of future self-continuity and its connection to financial outcomes such as savings behavior and global financial well-being. We also examine the strength of these associations in the face of a host of other relevant constructs. Overall, this research replicates and extends existing work on the relationship between future self-continuity and financial decision-making among a representative population, and sheds further insight on its potential implications for interventions aimed at enhancing long-term financial well-being.

Sustainable financial investments play a role in mitigating climate change. With this research, we explored how decision-makers use three investment strategies illustrating different primary motives: (1) money maximization (economic self-interest), (2) exclusion (not personally harming the environment), and (3) inclusion (helping the environment most efficiently). The relative use of these strategies was tested within a novel investment paradigm, aiming to artificially create a trade-off between financial payoffs (money maximization), environmental purity (exclusion), and environmental impact (inclusion). We recruited 1422 participants online and let them make ten consecutive investment choices, with incentivized outcomes, both monetary (i.e., potential bonus payment) and environmental (i.e., number of trees planted). We tested the change in investment choices using a between-subject design with four conditions, a control, and one default condition for each investment strategy. The three strategies were chosen about equally often in the control condition, but we find that this investment pattern could be altered by a default intervention. Preliminary evidence suggests that participants primarily using the exclusion and inclusion strategies differ in their moral reasoning. Utilitarianism better predicted the inclusion strategy, whereas high self-importance of moral identity better predicted the exclusion strategy.