Content uploaded by Jonathan Schooler
Author content
All content in this area was uploaded by Jonathan Schooler on Feb 15, 2014
Content may be subject to copyright.
Introspecting about Reasons can Reduce Post-Choice Satisfaction
Abstract
This study tested the prediction that introspecting about the reasons for one's preferences would reduce satisfaction with a consumer choice. Subjects evaluated two types of posters and then chose one to take home. Those instructed to think about their reasons chose a different type of poster than control subjects and, when contacted 3 weeks later, were less satisfied with their choice. When people think about reasons, they appear to focus on attributes of the stimulus that are easy to verbalize and seem like plausible reasons but may not be important causes of their initial evaluations. When these attributes imply a new evaluation of the stimulus, people change their attitudes and base their choices on these new attitudes. Over time, however, people's initial evaluation of the stimulus seems to return, and they come to regret choices based on the new attitudes.
... To test this, we design a full factorial experiment to systematically examine and disentangle the impact of multiple elements of prompt architecture. Specifically, in addition to order and labeling [32], we test two other classic choice architectural elements demonstrated in psychology-framing [33] and justification [34,35]-and examine how these elements interact with order and label bias within the same prompt. Mirroring findings in choice architecture, we find robust evidence that order, label, framing, and justification affect LLM responses. ...
... Third, to examine other elements of the prompt architecture, we vary the framing by asking GPT-4 either which set is closer to the first set or which set is farther away [33]. Finally, following work on justification [34,35], we vary whether we ask GPT-4 to justify its answer. ...
... However, research also finds that justification can bias human decision makers in other ways. For example, asking for justification leads people to select options that contain features that are easier to articulate, even if these options are less preferred (e.g., a humorous poster vs. an abstract, aesthetic one) [34]. Although LLMs do not experience emotions or motivation, structural patterns in the training set could lead to similar results. ...
We examine how the seemingly arbitrary way a prompt is posed, which we term “prompt architecture,” influences responses provided by large language models (LLMs). Five large-scale, full-factorial experiments performing standard (zero-shot) similarity evaluation tasks using GPT-3, GPT-4, and Llama 3.1 document how several features of prompt architecture (order, label, framing, and justification) interact to produce methodological artifacts, a form of statistical bias. We find robust evidence that these four elements unduly affect responses across all models, and although we observe differences between GPT-3 and GPT-4, the changes are not necessarily for the better. Specifically, LLMs demonstrate both response-order bias and label bias, and framing and justification moderate these biases. We then test different strategies intended to reduce methodological artifacts. Specifying to the LLM that the order and labels of items have been randomized does not alleviate either response-order or label bias, and the use of uncommon labels reduces (but does not eliminate) label bias but exacerbates response-order bias in GPT-4 (and does not reduce either bias in Llama 3.1). By contrast, aggregating across prompts generated using a full factorial design eliminates response-order and label bias. Overall, these findings highlight the inherent fallibility of any individual prompt when using LLMs, as any prompt contains characteristics that may subtly interact with a multitude of hidden associations embedded in rich language data.
... Deficiencies in these dimensions pose an obvious normative threat: if voters hold unstable, incoherent and extreme policy views, then their ability to hold politicians accountable is consequently diminished (see, for example, Achen and Bartels 2017, 306). The hopeful suggestion generated by the reason-as-causes model is that the quality of voters' attitudes might increase if only voters could be induced to 'think harder' about their political opinions, a suggestion buttressed by evidence from social psychology that shows greater introspection can indeed lead to more stable, more coherent, and less polarized attitudes in other domains (for example, Petty and Briñol 2011;Tesser 1978;Wilson et al. 1993;Wilson, Kraft and Dunn 1989;Dijksterhuis 2004). However, I show that increased cognitive effort does not result in such salutary effects for three important measures of political attitude quality (Price and Neijens, 1997). ...
... Existing evidence from social and cognitive psychology suggests that engaging in processes of reasoning can affect the attitudes people endorse (for example, Tesser 1978). In particular, introspecting about reasons appears to affect the decisions that people make and the satisfaction they subsequently feel from those decisions (Wilson, Kraft and Dunn 1989;Wilson and Schooler 1991;Wilson et al. 1993;Dijksterhuis 2004;Simonson 1989;Hsee 1999). The broad conclusion of this literature is that 'people who reason more act differently from those who reason less or not at all' (Mercier and Sperber 2018, 253). ...
... Finally, my results contrast with a well-established literature in social psychology which finds that asking people to explain the reasons for their attitudes can change the attitudes that they express (for example, Wilson, Kraft and Dunn 1989;Wilson and Schooler 1991;Wilson et al. 1993;Dijksterhuis 2004;Simonson 1989;Hsee 1999). In most cases, this research focuses on reason-giving in non-political settings, which provokes the question of whether there is something distinct about the process of reasoning about politics that prevents introspection from having the effects that are apparent elsewhere. ...
What are the effects of reason-giving on political attitudes? Both political philosophers and political scientists have speculated that defending proposals with reasons may change voters’ preferences. However, while models of attitude formation predict that the explicit justification of one’s political views may result in attitudes that are more ideologically consistent, less polarized, and more stable, empirical work has not assessed the connection between reason-giving and attitudes. Implementing a survey experiment in which some respondents provide reasons before stating their opinions on six issues in UK politics, I find that reason-giving has very limited effects on the constraint, stability, or polarization of the public’s political attitudes. These findings have important implications for our understanding of deliberative conceptions of democracy – in which reason-giving is a central component – as well as for our understanding of the quality of voters’ political opinions.
... In situations where there are reliable and valid cues in the environment for decision making and there is an opportunity to learn the outcomes of different scenarios through repeated practice in these environments, then valid intuitions can be learnt through experience (Hogarth, 2001;Kahneman & Klein, 2009;Klein, 2015). Second, in experimental studies where participants provide justification for their consumer choices, decision satisfaction is worse than when no justifications are required (Wilson et al., 1993). Generating reasons focuses attention on plausible but less important criteria which influence the decision that is made, to its detriment (Wilson & Schooler, 1991). ...
... Despite the evidence that System 1 processes can lead to biases in abstract decision-making problems (Frederick, 2005;Milkman et al., 2009), intuitive processes may allow individuals to more easily access subconscious information about choice options, including their own affective responses to the choices available (Pham et al., 2012). Relying on intuitive decision making processes may also prevent decision makers from being distracted by factors such as what they are meant to like about products, and instead focus on what automatically speaks to them or attracts them to certain options (Wilson & Schooler, 1991;Wilson et al., 1993). Intuition and heuristics for problem solving and decision making exist within human cognition, and while they are often argued to be a cause of mistakes under normative Procedure Participants first read the information sheet and consented to participate. ...
Which approach is better for making everyday decisions: intuitive or analytic thinking? Dual process theories of reasoning and decision-making propose that System 1 is intuitive, fast, and automatic, while System 2 is systematic, slow, and analytical. Previous research has highlighted the cognitive biases associated with System 1 and the rational performance of System 2. However, these processes have not been tested on real-life, everyday decisions to assess how accurately they predict satisfaction and happiness with the decision. In two experiments, participants made the choices of a meal to order and a movie to see and measured their use of System 1 or System 2 processes during the choice. They rated their satisfaction and happiness with the choice directly after deciding, and several days later after the decision outcome had been experienced. System 1 use was associated with increased satisfaction and happiness with the decision both immediately and after the decision outcome had been experienced, and also with how good the outcome of the decision was. System 2 use was largely not associated with satisfaction and happiness. System 1 intuitive thought is more effective for everyday choices if one’s goal is to be happy and satisfied.
... While this may be the case, it is also possible that people, when consulting God's guidance, end up making better choices by tapping into useful psychological processes. Relying on God for decisions could make people attuned to their intuitions, something that has been linked to improved decision making, at least in certain contexts (Wilson et al., 1993). Thinking about what God would want might also induce people to take a 3 rd person perspective, something that has been shown to improve judgment (Grossmann & Kross, 2014). ...
To confront difficult decisions, religious believers commonly seek guidance from sacred sources of knowledge, such as an all-knowing God. We predicted that engaging in this practice may be related to decision satisfaction. Specifically, across 3 studies (N= 2,474; two pre-registered), and three countries (U.S., Sweden, Canada), we explored whether religious believers who report consulting God during decisions experienced greater decision and life satisfaction, particularly when they also held a belief in God’s omniscience. Religious people who consulted God during decisions reported increased life satisfaction initially (Study 1a), and over time (Study 1b). Further, consistent with the proposed mechanism, when consulting God for guidance, participants who perceived God as especially omniscient (but not omnipotent or predictable) experienced greater decision satisfaction for past decisions (Study 2), anticipated future decisions (Study 2), and immediately following a decision (Study 3). Findings suggest that religious believers may benefit from seeking guidance from an omniscient God.
... Many studies report similar findings (Allman & Woodward 2008;Gigerenzer 2007;Gigerenzer & Gaissmaier 2011;Wilson 2002;Wilson et al. 1993;Woodward & Allman 2007), and they are all consistent with the idea that in some types of difficult decision-making, our automatic system consistently outperforms our manual system. But it's important to understand that this is not just due to the shortcomings of the manual system. ...
The longstanding view of doctors as scientists has been an emphasis in the MCAT and medical school training. However, the AAMC recommended recognizing the importance of social and behavioral science for medicine. There is also a growing realization that being a smart problem solver and the physician as scientist model emphasizes a cold cognitive problem-solving paradigm that overlooks other human capacities that may be critical to medical reasoning and decision-making. Considering a smart physician versus a wise physician, intelligence and problem-solving are important, but a wise physician can use other important capacities beyond intelligence and rationality. This could benefit patients by introducing patient and family perspective taking, as well as compassion in doctor-patient interaction. By reconceptualizing professions from the perspective of practical wisdom, this may increase resilience to problems such as burnout. I outline some psychological capacities viewed as important in wise reasoning that are not about traditional views of intelligence. I argue that wise reasoning is not a native talent but a skill that can be developed. I argue that different kinds of experiences can increase aspects of empathy, epistemic humility, perspective taking, and wise reasoning and I examine evidence that wise reasoning may increase resilience.
This research examines how a decision‐maker's perceived vulnerability influences their susceptibility to the “anecdotal bias,” a phenomenon where statistical evidence is disregarded in favor of anecdotal information. Across six studies, our research shows that high vulnerability aggravates the anecdotal bias instead of reducing it. Study 1 provides preliminary evidence that high vulnerability exacerbates the anecdotal bias among individuals seeking decision‐relevant information in the context of the COVID‐19 pandemic. Studies 2A and 2B demonstrate that high vulnerability intensifies the anecdotal bias in different decision contexts. Study 3 replicates these findings and identifies negative emotional arousal as a key mechanism underlying this effect. Study 4 examines the moderating role of personal relevance, showing that when individuals make decisions for others (vs. themselves), high vulnerability does not lead to the anecdotal bias. Moreover, it is cognitive disruption and intuitive thinking caused by negative emotional arousal that increases reliance on anecdotal (vs. statistical) information. Finally, Study 5 demonstrates the moderating effect of mindfulness meditation, highlighting its role as a preemptive safeguard against this biased behavior. Theoretical contributions and practical implications of these findings are discussed.
Background
Evidence is lacking on the most effective values clarification methods (VCMs) in patient decision aids (PtDAs). We tested the effects of an adaptive conjoint analysis (ACA)–based VCM compared with a ranking-based VCM and no VCM on several decision-related outcomes, with the decisional conflict and its subscale “perceived values clarity” as primary outcomes.
Design
Online experimental study with 3 conditions: no VCM versus ranking-based VCM versus ACA-based VCM ( N = 282; M age = 63.11 y, s = 12.12), with the latter 2 conditions including attributes important for a lung cancer treatment decision. We assessed 1) decisional conflict, 2) perceived values clarity (decisional conflict subscale), 3) perceived cognitive load, 4) anticipated regret, 5) ambivalence, 6) preparedness for decision making, 7) hypothetical treatment preference, and 8) values congruence (proxy). We performed analysis of covariance and linear regression. Age and level of deliberation were included as potential moderators, and we controlled for subjective numeracy (covariate). We exploratively tested the moderating effects of subjective numeracy and health literacy (without covariates).
Results
We found no significant effect of type of VCM on overall decisional conflict or perceived values clarity. Age had a moderating effect: in younger participants, no VCM (v. ranking-based VCM) led to more values clarity, while in older participants, a ranking-based VCM (v. no VCM) led to more values clarity. Completing the ACA-based VCM, compared with no VCM, resulted in more values congruence.
Limitations
The hypothetical choice situation might have induced lower levels of cognitive/affective involvement in the decision.
Conclusions
This study found mixed effects of an ACA-based VCM. It did not decrease decisional conflict or increase perceived values clarity, yet it did improve values congruence.
Implications
Completion of an ACA-based VCM in a PtDA may increase values congruence.
Highlights
An adaptive conjoint analysis or a ranking-based values clarification method did not decrease analog patients’ decisional conflict nor did it increase their perceived values clarity. In younger participants, no VCM (v. ranking-based VCM) led to more values clarity, while in older participants, a ranking-based VCM (v. no VCM) led to more values clarity. An adaptive conjoint analysis task for values clarification resulted in more values congruence.
The goal of the study was to determine which aspects of interpersonal touch interactions lead to a positive or negative experience. Previous research has focused primarily on physical characteristics. We suggest that this may not be sufficient to fully capture the complexity of the experience. Specifically, we examined how fulfilment of psychological needs influences touch experiences and how this relates to physical touch characteristics and situational factors.In two mixed-method studies, participants described their most positive and most negative interpersonal touch experience within a specific time frame. They reported fulfilment of nine needs, affect, intention, and reason for positivity/negativity, as well as the body part(s) touched, location, type of touch, interaction partner, and particular touch characteristics (e.g. humidity).Positive and negative touch experiences shared similar touch types, locations, and body parts touched, but differed in intended purpose and reasons. Overall, the valence of a touch experience could be predicted from fulfilment of relatedness, the interaction partner and initiator, and physical touch characteristics. Positive affect increased with need fulfilment, and negative affect decreased.The results highlight the importance of relatedness and reciprocity for the valence of touch, and emphasise the need to incorporate psychological needs in touch research.
The effects of different types of responses to a depressed mood on the duration and severity of the mood were examined. On the basis of Nolen-Hoeksema's (1987) response styles theory of depression, it was hypothesized that distracting, active responses would be more effective in alleviating a depressed mood than would ruminative, passive responses. A depressed mood was induced in 35 male and 34 female Ss, and subjects were randomly assigned to engage in 1 of 4 types of responses: an active task that distracted them from their mood; a passive, distracting task; an active task designed to lead to ruminations about their mood; or a passive, ruminative task. As predicted, the greatest remediation of depressed mood was found in Ss in the distracting–active response condition, followed in order by the distracting–passive, ruminative–active, and ruminative–passive response conditions. Degree of rumination had a greater impact on remediation of depressive affect than level of activity, with greater rumination leading to lesser remediation of depressive affect. In addition, the effects of the response tasks were limited to depressed mood. The implications of these results for interventions with depressed persons are discussed.
This chapter provides history of the ape language research, indicating past and current work. It raises some relevant theoretical issues and relates the research to investigations or issues in the fields of experimental and developmental psychology, linguistic analysis, anthropology, and ethology. The ape language projects are best considered in their relation to the issue of mental continuity between man and animal. They reveal a potential bridge between ape abilities that may be precursors to man's linguistic and cognitive abilities. Communicative media utilized in the projects are hand signing. The signing projects, pioneered by the Gardners, have the advantage that the full repertoire of vocabulary items is available to the apes at all times. Therefore, the apes have more opportunity for creative productions to describe and name objects and events in the environment and, potentially, to create grammatically novel sentences. The signing projects share a number of methodological and interpretative problems. The chapter discusses some of the difficulties related to the process of acquiring data, such as the use of daily diaries, the procedures for conducting reliability tests, and anecdotal reporting of novel word use and combinations.
Tested the attribution theory notion that making the extrinsic rewards that individuals provide for their partners salient would decrease the amount of love and, perhaps, liking that the partners actually report for their girlfriends or boyfriends. 19 undergraduate dating couples were induced to adopt either an extrinsic or an intrinsic cognitive set regarding their reasons for associating with their partners. Then their liking and love for each other were assessed. Results indicate that the induced salience of extrinsic rewards led couples to report less love but not less liking. (28 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Examined the distinction between (a) classical views of attitudes as stable dispositions based on beliefs and prior experiences accessed from memory and (b) the self-perception analysis of attitudes as relatively transitory responses dependent on current contextual cues. The access of 159 undergraduates to relevant information in memory was assessed by the number of issue-related beliefs and prior experiences they could retrieve. In Exp I, Ss read a counter-attitudinal message. As expected, Ss with little access to beliefs and prior experiences, in comparison to those with greater access, changed their opinions to be more consistent with the message position. In Exp II, Ss agreed to advocate a pro-attitudinal position for a reward or no reward. As expected, Ss with access to little relevant information in memory inferred their attitudes from their decision to proselytize: Rewarded Ss explained their decision less in terms of their own belief in the position than did Ss who had greater access to information in favor of the position, who tended to attribute their decision to a belief in the topic, regardless of the presence or absence of the reward. (34 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Two experiments with 268 paid volunteers investigated the possibility that assessment of confidence is biased by attempts to justify one's chosen answer. These attempts include selectively focusing on evidence supporting the chosen answer and disregarding evidence contradicting it. Exp I presented Ss with 2-alternative questions and required them to list reasons for and against each of the alternatives prior to choosing an answer and assessing the probability of its being correct. This procedure produced a marked improvement in the appropriateness of confidence judgments. Exp II simplified the manipulation by asking Ss first to choose an answer and then to list (a) 1 reason supporting that choice, (b) 1 reason contradicting it, or (c) 1 reason supporting and 1 reason contradicting. Only the listing of contradicting reasons improved the appropriateness of confidence. Correlational analyses of the data of Exp I strongly suggested that the confidence depends on the amount and strength of the evidence supporting the answer chosen. (21 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Previous research has produced contradictory findings regarding the effects of thought on the attitude–behavior relation. In an attempt to integrate these findings, the present authors proposed that thought may make either the affective or cognitive component of the attitude more salient and, thus, more important in the formation of the general evaluation and that behaviors may either be cognitively or affectively driven. It was hypothesized that a match between the attitude component emphasized by thought and the attitude component that drives behavior would increase the attitude–behavior relation, and alternatively, that a mismatch between the attitude components would decrease the relation. To test this hypothesis, 68 undergraduates focused either on the affective or cognitive component of their attitude before evaluating puzzles. The context of puzzle-playing behavior was varied to make the behavior either instrumental (cognitively driven) or consummatory (affectively driven), and Ss were given the opportunity to play with the puzzles. Correlations between the evaluations of the puzzles and the amount of time spent playing with the puzzles supported the hypothesis: Thought emphasizing the affective component produced evaluations that predicted consummatory behavior (affectively driven behaviors) but not instrumental behavior (cognitively driven). Alternatively, thought emphasizing the cognitive component produced evaluations that predicted instrumental behavior (cognitively driven) but not consummatory behavior. (20 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Reviews evidence which suggests that there may be little or no direct introspective access to higher order cognitive processes. Ss are sometimes (a) unaware of the existence of a stimulus that importantly influenced a response, (b) unaware of the existence of the response, and (c) unaware that the stimulus has affected the response. It is proposed that when people attempt to report on their cognitive processes, that is, on the processes mediating the effects of a stimulus on a response, they do not do so on the basis of any true introspection. Instead, their reports are based on a priori, implicit causal theories, or judgments about the extent to which a particular stimulus is a plausible cause of a given response. This suggests that though people may not be able to observe directly their cognitive processes, they will sometimes be able to report accurately about them. Accurate reports will occur when influential stimuli are salient and are plausible causes of the responses they produce, and will not occur when stimuli are not salient or are not plausible causes. (86 ref)
Recent social psychological research has indicated that merely thinking about some object will, under certain conditions, result in more intensified feelings (attitude polarization) toward that object. Implications of a selfgenerated attitude change model were tested for therapeutic interventions concerning phobias, particularly fear of public speaking. Subjects, who had previously indicated that they had a strong fear of public speaking, were exposed to one of three treatment conditions: Process Constraints condition, in which subjects were asked to make their beliefs and the derivations of those beliefs about public speaking explicit; Catharsis condition, in which subjects thought in an unconstrained manner about public speaking; and a Control condition. Relative to control subjects, subjects in the Process Constraints condition were expected to show reduced affect, whereas those in the Catharsis condition were expected to show increased affect. Results supported these hypotheses on self-report and physiological measures, but not on behavioral ratings. Implications of these results for therapeutic intervention, particularly Ellis's RET, were discussed.
The chapter presents evidence consistent with the observations of Roethke and Vargas Llosa that introspection can be disruptive. The focus is on one type of introspection-thinking—the reasons for one's feelings. The chapter demonstrates that this type of thought can cause people to change their minds about the way they feel and lead to a disconnection between their attitudes and their behavior. It is clear that asking people to think about reasons will often produce attitude change, particularly for affectively based attitudes. The direction of this change, however, has been difficult to predict. The chapter explains people who think about reasons and end up with an attitude that is significantly more negative or positive, on the average, than the attitudes of control subjects. The direction of attitude change is difficult to predict, because it is closely related to the hypothesis about the generation of a biased sample of reasons. In the chapter, there are at least two sorts of harmful attitudes that might be changed by thinking about reasons—those that are undesirable from the individual's perspective and those that are undesirable from a societal perspective.
To test Bem's (1972) hypothesis that individuals infer their attitudes from information about their behavior in a given context, subjects were asked their attitudes about a course after reporting their behaviors for the course and receiving a linguistic manipulation intended to produce an intrinsic or extrinsic cognitive context (set). Intrinsic set was manipulated by having subjects complete statements of the form “I generally do X because I …”. Extrinsic set was manipulated by having subjects complete statements of the form “I generally do X in order to …”. A pretest found that the intrinsic form resulted in a class perceived as more interesting than useful while the extrinsic form resulted in a class perceived as more useful than interesting. The present experiment found that subjects' attitudes are derived from information about different behaviors under the two cognitive sets. The attitudes of extrinsic set subjects were correlated with behaviors relevant to obtaining course out-comes like grades (contingent behaviors) and uncorrelated with behaviors relevant to personal interest (noncontingent behaviors); the opposite was found for intrinsic set subjects.