We report on three pre-registered studies testing whether people in the positionof describing a decision problem to decision-makers exploit this opportunity fortheir benefit, by choosing descriptions that may be potentially beneficial forthemselves. In Study 1, recipients of an extreme dictator game (where dictatorscan either take the whole pie for themselves or give it entirely to thereceiver) are asked to choose the instructions used to introduce the game todictators, from six different instructions known from previous research toaffect dictators’ decisions. The results demonstrate that some dictatorgame recipients tend to choose instructions that make them more likely toreceive a higher payoff. Study 2 shows that people who choose descriptions thatmake them more likely to receive a higher payoff indeed believe that they willreceive a higher payoff. Study 3 shows that receivers are more likely thandictators to choose these self-serving descriptions. In sum, our work suggeststhat some people choose descriptions that are beneficial to themselves; we alsofound some evidence that deliberative thinking and young age are associated withthis tendency.

Loss aversion, the argument that losses are given more weight than gains, has been recently shown to be absent in small losses. However, a series of studies by Mrkva et al. (2020) appear to demonstrate the existence of loss aversion even for smaller losses. We re-ran Mrkva et al.’s decision tasks after removing features of the task that differentiated losses from the gains, particularly asymmetries in sizes of gains and losses, an increasing order of losses, and status quo effects. The results show that we replicate Mrkva et al.’s (2020) findings in their original paradigm with online participants, yet in five studies where gains and losses were symmetrically presented in random order (n = 2,001), we find no loss aversion for small amounts, with loss aversion surfacing very weakly only for average losses of $40 (mean λ = 1.16). We do find loss aversion for higher amounts such as $100 (mean λ = 1.54) though it is not as extreme as previously reported. Furthermore, we find weak correlation between the endowment effect and loss aversion, with the former effect existing simultaneously with no loss aversion. Thus, when items are presented symmetrically, significant loss aversion emerges only for large losses, suggesting that it cannot be argued that (all) “losses loom larger than gains.”

How do people make everyday decisions in order to achieve the most successful outcome? Decision making research typically evaluates choices according to their expected utility. However, this research largely focuses on abstract or hypothetical tasks and rarely investigates whether the outcome is successful and satisfying for the decision maker. Instead, we use an everyday decision making task in which participants describe a personally meaningful decision they are currently facing. We investigate the decision processes used to make this decision, and evaluate how successful and satisfying the outcome of the decision is for them. We examine how well analytic, attribute-based processes explain everyday decision making and predict decision outcomes, and we compare these processes to associative processes elicited through free association. We also examine the characteristics of decisions and individuals that are associated with good decision outcomes. Across three experiments we found that: 1) an analytic decision analysis of everyday decisions is not superior to simpler attribute-based processes in predicting decision outcomes; 2) contrary to research linking associative cognition to biases, free association generates valid cues that predict choice and decision outcomes as effectively as attribute-based approaches; 3) contrary to research favouring either attribute-based or associative processes, combining both attribute-based and associates best explains everyday decisions and most accurately predicts decision outcomes; and 4) individuals with a tendency to attempt analytic thinking do not make more successful everyday decisions. Instead, frequency, simplicity, and knowledge of the decision predict success. We propose that attribute-based and associative processes, in combination, both explain everyday decision making and predict successful decision outcomes.

When evaluating a charity by itself, people tend to overweight overhead costs inrelation to cost-effectiveness. However, when evaluating charities side by side,they base their donations on cost-effectiveness. I conducted a replication andextension of Caviola et al. (2014; Study 1) using a 3 (HighOverhead/Effectiveness, Low Overhead/Effectiveness, Both) x 2 (Humans, Animals)between-subjects design. I found that the overhead ratio is an easier attributeto evaluate than cost-effectiveness in separate evaluation, and, in jointevaluation, people allocate donations based on cost-effectiveness. This effectwas observed for human charities, and to a lesser extent, for animalcharities.

Previous research has demonstrated systematic discrepancies between description-and experience-based risky choices. This description-experiencegap has been attributed to several factors such as reliance onsmall samples and differential probability weighting patterns. Becausecontext-dependent outcome expectation regarding safe options might influenceexperience-based risky choices, it constitutes another potential contributor tothe gap. Using a free-sampling paradigm and risky options with rare outcomesthat were either attractive or unattractive relative to the frequent ones, twoexperiments examined the existence and impact of such outcome expectation inexperience-based risky choices. Both experiments had two information conditions:hint information meant to eliminate outcome expectation was provided in onecondition but not the other. Experiment 1, which indicated the numbers ofpossible outcomes regarding both safe and risky options under the hintcondition, revealed different choice behaviors regarding risky-safe trialsbetween the two information conditions, no matter whether the rare outcome ofthe risky option in such a trial (i.e., the local context) wasattractive or unattractive. However, this result provided only indirect evidencefor the role of outcome expectation because it was unclear whether the hintinformation affected only the outcome expectation and thus evaluation of safeoptions or the evaluations of both safe and risky ones. With refined hintinformation arguably removing potential impacts on the evaluation of riskyoptions, Experiment 2 showed that expectation of a non-existent rare outcome ofsafe options did contribute to experience-based risky choices. In addition, itappeared that the rare outcomes of the risky options in other decision problemspresented in the same experiment (i.e., the global context)also affected outcome expectation. Future research could investigate how theinteraction between local and global contexts determines outcome expectation todeepen our understanding of its contribution to experience-based risky choiceand the description-experience gap.

To what degree do people prefer to choose for themselves and what drives this preference? Is it memory-based and results from a life-long association between choices and better outcomes, or is the process of choice itself reinforcing? In a new paradigm, across 6 experiments, participants experienced both ’Own Choice’ and ’Computer Picks’ conditions with identical outcomes before selecting which condition to re-experience in the final part of the experiment. Consistent with previous work, an overwhelming majority ( 83%) preferred own-choice. Several variations of the paradigm reveal that (1) Preference For Choice (PFC) is reduced when thinking about the task without actually choosing in it, (2) PFC is substantially reduced by choice-unrelated cognitive load, and (3) Preference For Choice is further diminished when selection is based on criteria other than one’s preferences. Across experiments, participants’ self-rated enjoyment predicted a significant portion of their PFC, while their perceived gains had little to no predictive value. If PFC stems solely from past reinforcement learning (i.e., memory) then neither performing another few scores of choices nor adding cognitive load to that sequence of choices would be expected to dramatically affect it. Hence, our findings suggest that a significant part of this preference stems from the process of choice itself, and that the experience it confers can itself be reinforcing. We discuss the implications of the proposed mechanism for PFC, which leads us to the prediction that PFC may be muted or even reversed under specific conditions and what this means for when the ‘opposite’ effect – sticking with the default – will occur.

Fallacies are a particular type of informal argument that are psychologicallycompelling and often used for rhetorical purposes. Fallacies are unreasonablebecause the reasons they provide for their claims are irrelevant orinsufficient. Ability to recognize the weakness of fallacies is part of what wecall argument literacy and imporatant in rational thinking. Here we examineclassic fallacies of types found in textbooks. In an experiment, participantsevaluated the quality of fallacies and reasonable arguments. We instructedparticipants to think either intuitively, using their first impressions, oranalytically, using rational deliberation. We analyzed responses, responsetimes, and cursor trajectories (captured using mouse tracking). The resultsindicate that instructions to think analytically made people spend more time onthe task but did not make them change their minds more often. When participantsmade errors, they were drawn towards the correct response, while respondingcorrectly was more straightforward. The results are compatible with“smart intuition” accounts of dual-process theories of reasoning,rather than with corrective default-interventionist accounts. The findings arediscussed in relation to whether theories developed to account for formalreasoning can help to explain the processing of everyday arguments.

Experience is the best teacher. Yet, in the context of repeated decisions, experience was found to trigger deviations from maximization in the direction of underweighting of rare events. Evaluations of alternative explanations for this bias led to contradicting conclusions. Studies that focused on the aggregate choice rates, including a series of choice prediction competitions, favored the assumption that this bias reflects reliance on small samples. In contrast, studies that focused on individual decisions suggest that the bias reflects a strong myopic tendency by a significant minority of participants. The current analysis clarifies the apparent inconsistency by reanalyzing a data set that previously led to contradicting conclusions. Our analysis suggests that the apparent inconsistency reflects the differing focus of the cognitive models. Specifically, sequential adjustment models (that assume sensitivity to the payoffs’ weighted averages) tend to find support for the hypothesis that the deviations from maximization are a product of strong positive recency (a form of myopia). Conversely, models assuming random sampling of past experiences tend to find support to the hypothesis that the deviations reflect reliance on small samples. We propose that the debate should be resolved by using the assumptions that provide better predictions. Applying this solution to the data set we analyzed shows that the random sampling assumption outperforms the weighted average assumption both when predicting the aggregate choice rates and when predicting the individual decisions.

Granting a short-term loan is a critical decision. A great deal of research hasconcerned the prediction of credit default, notably through Machine Learning(ML) algorithms. However, given that their black-box nature has sometimes led tounwanted outcomes, comprehensibility in ML guided decision-making strategies hasbecome more important. In many domains, transparency and accountability are nolonger optional. In this article, instead of opposing white-box againstblack-box models, we use a multi-step procedure that combines the Fast andFrugal Tree (FFT) methodology of Martignon et al. (2005) and Phillips et al.(2017) with the extraction of post-hoc explainable informationfrom ensemble ML models. New interpretable models are then built thanks to theinclusion of explainable ML outputs chosen by human intervention. Ourmethodology improves significantly the accuracy of the FFT predictions whilepreserving their explainable nature. We apply our approach to a dataset ofshort-term loans granted to borrowers in the UK, and show how complex machinelearning can challenge simpler machines and help decision makers.

It is a widely held view that people rely on incomplete information to find a relationship partner, resulting in non-compensatory choice heuristics. However, recent experimental work typically finds that partner choice follows compensatory choice strategies. To bridge this gap between theory and experimental evidence, we characterize the mate choice problem by distinguishing the information search process from the evaluation process. In an eye-tracking experiment and a MouseLab experiment, we show that people display strong value-directed search heuristics in response to all types of cues and that the magnitude of value-directed searches increases with cue primacy. Cue primacy also explains the interaction effect of cue type and participant sex on the extent of valued-directed search. We further argue that value-directed searching does not necessarily lead to non-compensatory choice rules but may serve compensatory decision-making. Our results demonstrate that people may adopt remarkably smart search heuristics to find an ideal partner efficiently.

Believers of karma believe in ethical causation where good and bad outcomes canbe traced to past moral and immoral acts. Karmic belief may have importantinterpersonal consequences. We investigated whether American Christians expectmore trustworthiness from (and are more likely to trust) interaction partnerswho believe in karma. We conducted an incentivized study of the trust game whereinteraction partners had different beliefs in karma and God. Participantsexpected more trustworthiness from (and were more likely to trust) karmabelievers. Expectations did not match actual behavior: karmic belief was notassociated with actual trustworthiness. These findings suggest that people mayuse others’ karmic belief as a cue to predict their trustworthiness butwould err when doing so.

We examine the effect of an irrelevant task that may become a reference point on subjects’ effort, feelings and perceptions. All subjects complete up to 25 tasks and are paid $0.10 per task solved correctly. However, some subjects have an easy task of finding one letter and others have a hard task of finding two letters. In the irrelevant-task treatment conditions subjects are told about the two types of tasks and are then assigned randomly to one. In addition, there are two control conditions, and in each control condition subjects are assigned to a specific task without the other task being possible or mentioned. Subjects in the irrelevant-task treatments express more positive (negative) feelings when assigned to the easy (hard) task. The control conditions that have no reference point of another task are in between the two irrelevant-task treatments in the feeling ratings. We hypothesized that for a given task, the subjects in the experimental conditions that have more positive feelings will also solve more tasks, but this hypothesis was not supported by the data. Finally, subjects who receive the easy task complete more tasks than the ones with the hard task.

The accuracy of human forecasters is often reduced because of incompleteinformation and cognitive biases that affect the judges. One approach to improvethe accuracy of the forecasts is to recalibrate them by means of non-lineartransformations that are sensitive to the direction and the magnitude of thebiases. Previous work on recalibration has focused on binary forecasts. Wepropose an extension of this approach by developing an algorithm that uses asingle free parameter to recalibrate complete subjective probabilitydistributions. We illustrate the approach with data from the quarterly Survey ofProfessional Forecasters (SPF) conducted by the European Central Bank (ECB),document the potential benefits of this approach, and show how it can be used inpractical applications.

Existing evidence from laboratory experiments finds that a match is likely to increase charitable donations by more than a theoretically equivalent rebate. A number of explanations have been proposed for this in the literature. One idea, which has never been tested, is that people consider a match to be more generous, because unlike the rebate, there is no reward for making a donation in the match setting. We design a survey to determine whether people do consider matches more generous than rebates, and probe the reasons subjects give for their answers. We find that a significant number of people do consider rebates less generous because of the reward associated with donations in such a setting.

Machines are increasingly used to make decisions. We investigated people’sbeliefs about how they do so. In six experiments, participants (totalN = 2664) predicted how computer and human judges woulddecide legal cases on the basis of limited evidence — eitherindividuating information from witness testimony or base-rate information. InExperiments 1 to 4, participants predicted that computer judges would be morelikely than human ones to reach a guilty verdict, regardless of which kind ofevidence was available. Besides asking about punishment, Experiment 5 alsoincluded conditions where the judge had to decide whether to reward suspectedhelpful behavior. Participants again predicted that computer judges would bemore likely than human judges to decide based on the available evidence, butalso predicted that computer judges would be relatively more punitive than humanones. Also, whereas participants predicted the human judge would give moreweight to individuating than base-rate evidence, they expected the computerjudge to be insensitive to the distinction between these kinds of evidence.Finally, Experiment 6 replicated the finding that people expect greatersensitivity to the distinction between individuating and base-rate informationfrom humans than computers, but found that the use of cartoon images, as in thefirst four studies, prevented this effect. Overall, the findings suggest peopleexpect machines to differ from humans in how they weigh different kinds ofinformation when deciding.

This study compares Chinese people’s trust and trustworthiness, risk attitude, and time preference before and after the onset of the COVID-19 pandemic in China. We compare the preferences of subjects in two online experiments with samples drawn from 31 provinces across mainland China before and after the onset of the pandemic. We test two competing hypotheses regarding trust and trustworthiness. On the one hand, the outbreak as a collective threat could enhance in-group cohesion and cooperation and thus increase trust and trustworthiness. On the other hand, to the extent that people expect their future income to decline, they may become more self-protective and self-controlled, and thus less trusting and trustworthy and more risk averse and patient. Comparing before and after the onset, we found that the subjects increased in trustworthiness. After the onset, trust and trustworthiness (and risk aversion and present bias too) were positively correlated with the COVID-19 prevalence rate in the provinces. Subjects with more pessimistic expectations about income change showed more risk aversion and lower discount rates, supporting the speculation concerning self-control.

In two experiments, participants decided on each of several trials whether or notto take a risk. If they chose to take the risk, they had a relatively highprobability (from 75% to 95%) of winning a small number of points and arelatively low probability (5% to 25%) of losing a large number of points. Theloss amounts varied so that the expected value of taking the risk was positiveon some trials, zero on others, and negative on the rest. The main independentvariable was whether the probability of losing was communicated using numericalpercentages or icon arrays. Both experiments included random icon arrays, inwhich the icons representing losses were randomly distributed throughout thearray. Experiment 2 also included grouped icon arrays, in which the iconsrepresenting losses were grouped at the bottom of the array. Neither type oficon array led to better performance in the task. However, the random iconarrays led to less risk taking than the numerical percentages or the groupedicon arrays, especially at the higher loss probabilities. In a third experiment,participants made direct judgments of the percentages and probabilitiesrepresented by the icon arrays from Experiment 2. The results supported the ideathat random arrays lead to less risk taking because they are perceived torepresent greater loss probabilities. These results have several implicationsfor the study of icon arrays and their use in risk communication.

In a controlled laboratory experiment we investigate whether time pressure influences voting decisions, and in particular the degree of strategic (insincere) voting. We find that participants under time constraints are more sincere when using the widely-employed Plurality Voting method. That is, time pressure might reduce strategic voting and hence misrepresentation of preferences. However, there are no effects for Approval Voting, in line with arguments that this method provides no incentives for strategic voting.

We propose the concept of ‘scientific contagion’ — a mental heuristic through which any form of scientific treatment transfers some essence of ‘science’ to the processed substance, thereby affecting its nature and social acceptability. This was tested regarding the potential treatment of water from natural sources before it is used for religious purposes, as many such sources have dangerous pollutants. For an ancient natural well having a religious narrative, most participants judged that the acceptability of water would be reduced for religious purposes but not for drinking if local officials scientifically treat the water. That is not the case if religious rituals are conducted on the water instead (Study 1). If water from a “holy river” is processed scientifically, most participants predicted that it would reduce acceptability for religious use while increasing acceptability for drinking (Study 2). Potential scientific treatment without altering the composition of water from a natural spring also decreased acceptability for religious use but there was no effect on acceptability for drinking or on willingness to pay money for the water (Study 3). A follow-up study comparing acceptability for different kinds of water sources — from a holy well, natural spring, and household tap water sourced from either underground wells or rivers found lower acceptability for religious usage compared to drinking after potential scientific treatment for all these waters, but more so for holy and natural waters (Study 4). These studies establish the phenomena of scientific contagion that could have significant social implications and open future directions.

Survey respondents over-forecast inflation: they expect it to be higher than itturns out to be. Furthermore, people are generally overconfident in theirforecasts. In two experiments, we show that providing outcome feedback thatinforms people of the actual level of the inflation that they have forecastreduces both over-forecasting and overconfidence in forecasts. Theseimprovements were preserved even after feedback had been withdrawn, a findingthat indicates that they were not produced because feedback had a temporaryincentive effect but because it had a more permanent learning effect. However,providing forecasters with more outcome feedback did not have a greater effect.Feedback appears to provide people with information about biases in theirjudgments and, once they have received that information, no additional advantageis obtained by giving it to them again. Reducing over-forecasting also had noclear effect on overall error. This was because providing outcome feedback afterevery judgment also affected the noise or random error in forecasts, increasingit by a sufficient amount to cancel out the benefits provided by the reductionin over-forecasting.