PDF(932 KB)
PDF(932 KB)
PDF(932 KB)
谁更听信AI?个体依恋风格对人工智能建议采纳的影响*
Who Listens to AI? The Role of Attachment Styles in Adopting Artificial Intelligence Advice
人工智能（AI）正日益成为个体获取建议的重要来源之一。依恋风格作为一种反映社会行为倾向的心理特质，塑造了个体对建议者相对稳定的预期与认知，为阐释AI建议采纳的个体差异提供了有力的理论框架。研究基于Bowlby的依恋理论，探究了个体依恋风格与建议来源（AI vs. 人类）对建议采纳的影响。两项实验研究均发现：（1）与人类建议相比，高依恋焦虑者更倾向于采纳AI的建议，即随着依恋焦虑程度的提升，个体对AI建议的采纳程度显著增加，但对人类建议的采纳程度没有显著变化；（2）高依恋回避者则未表现出对AI或人类建议的显著偏好。结果说明，依恋风格可以有效解释AI建议采纳的个体差异，依恋焦虑和依恋回避具有差异化的作用机制。
In the era of intelligence, artificial intelligence (AI) and humans have become two important sources of advice. Attachment style, a classic social trait in psychology, reflects an individual's stable mental representation of an advisor and their positive or negative evaluations. It provides an effective theoretical framework for explaining individuals’ inherent preferences for advice from humans or AI. Based on Bowlby's attachment theory, current research aims to explore how attachment styles and the source of advice (AI or humans) jointly influence individuals' advice adoption.
Two experiments were conducted to examine the interaction between attachment styles and advice source on advice-taking behavior. Experiment 1 utilized a 2 (attachment style: attachment anxiety, attachment avoidance) × 2 (advice source: human, AI) between-subjects design. A total of 196 participants were recruited through an online survey platform. The primary dependent variable was the weight of advice (WOA), measured using the Judge-Advisor System (JAS) paradigm, which quantifies the extent to which participants adjust their initial judgments based on the advice received. Attachment styles were assessed using the Chinese version of the Experiences in Close Relationships Scale (ECR). Control variables included gender, age, education level, and self-reported AI knowledge. Participants were asked to estimate the weight of a person in a photograph (initial assessment) and then received either human or AI-generated advice on the weight. They were subsequently asked to provide a final weight estimate (post-assessment). The difference between the initial and post-assessment estimates, relative to the advice received, was used to calculate the WOA. Experiment 2 employed an attachment priming paradigm to replicate and extend the findings from Experiment 1. A total of 248 participants were randomly assigned to one of three attachment priming conditions (secure attachment, attachment anxiety, attachment avoidance) and then completed the same advice-taking task as in Experiment 1.
The research reveals that (1) Compared with human advice, individuals with high attachment anxiety are more likely to adopt advice from AI. That is, as the level of attachment anxiety increases, individuals' adoption of AI advice significantly rises, while their adoption of human advice does not change significantly; (2) Individuals with high attachment avoidance do not show a significant preference for either AI advice or human advice. Individuals with high attachment anxiety may find AI advice more appealing due to its perceived objectivity and lack of social threat, which aligns with their desire for support without the fear of social rejection. In contrast, attachment avoidance does not appear to drive a clear preference for either AI or human advice, suggesting that the underlying mechanisms influencing advice acceptance may differ between attachment anxiety and avoidance. Future research should explore additional mediators and moderators that may explain these differential effects and consider the impact of attachment styles on advice-taking in more complex and ecologically valid decision-making scenarios.
In terms of theoretical contributions, this study introduces the attachment theory into the research on advice interaction between humans and artificial intelligence. It enriches the theoretical research findings on human-Machine interaction and provides a novel theoretical framework for understanding the contradictory phenomena of algorithm appreciation and algorithm aversion in the intelligent era. In terms of practical implications, the findings of this study provide insights for the design of AI-based decision support systems, especially in contexts where user trust and acceptance are crucial for effective human-Machine collaboration. For individual advice adoption, the study offers a reference for individuals to optimize their advice adoption and decision-making patterns in the intelligent era. Specifically, those who can rationally analyze the capability boundaries between AI and humans, and comprehensively evaluate advice from different sources, are more likely to make effective decisions. Individuals with high attachment anxiety may exhibit irrational algorithm appreciation, which could lead to the adoption of erroneous suggestions. This, in turn, may distort and negatively impact their cognition, necessitating attention to the potential risk of over-reliance.
建议采纳 / 依恋风格 / 人机协同 / 算法欣赏 / 算法厌恶
advice taking / attachment styles / human-machine collaboration / algorithm appreciation / algorithm aversion
| [1] |
杜秀芳, 胡卫红, 王静, 李方. (2022). 情境公开性与印象管理动机对建议采纳的影响. 心理科学, 45(5), 1198-1205.
|
| [2] |
段锦云, 古晓花, 孙露莹. (2016). 外显自尊、内隐自尊及其分离对建议采纳的影响. 心理学报, 48(4), 371-384.
|
| [3] |
李彩娜, 孙颖, 拓瑞, 刘佳. (2016). 安全依恋对人际信任的影响:依恋焦虑的调节效应. 心理学报, 48(8), 989-1001.
|
| [4] |
李同归, 加藤和生. (2006). 成人依恋的测量:亲密关系经历量表(ECR)中文版. 心理学报, 38(3), 399-406.
|
| [5] |
罗希彤, 潘亚峰. (2023). 人际视角下的建议互动:决策、社会认知过程与计算神经机制. 科学通报, 68(Z2), 3809-3822.
|
| [6] |
申琦, 王璐瑜. (2021). 当“机器人”成为社会行动者:人机交互关系中的刻板印象. 新闻与传播研究, 28(2), 37-52+127.
|
| [7] |
王林辉, 胡晟明, 董直庆. (2022). 人工智能技术、任务属性与职业可替代风险:来自微观层面的经验证据. 管理世界, 38(7), 60-79.
|
| [8] |
温忠麟, 侯杰泰, 张雷. (2005). 调节效应与中介效应的比较和应用. 心理学报, 37(2), 268-274.
|
| [9] |
乌尔里希·贝克, 王武龙. (2003). 从工业社会到风险社会(上篇)——关于人类生存、社会结构和生态启蒙等问题的思考. 马克思主义与现实, 3, 26-45.
|
| [10] |
吴新慧. (2020). 数字信任与数字社会信任重构. 学习与实践, 10, 87-96.
|
| [11] |
徐惊蛰, 谢晓非. (2009). 决策过程中的建议采纳. 心理科学进展, 17(5), 1016-1025.
|
| [12] |
张艳梅, 杜秀芳, 王修欣. (2015). 焦虑、建议者善意程度对个体建议采纳的影响. 心理科学, 38(5), 1155-1161.
|
| [13] |
周浩, 龙立荣. (2004). 共同方法偏差的统计检验与控制方法. 心理科学进展, 12(6), 942-950.
共同方法偏差是心理学研究中一个越来越被关注的课题，但国内心理学界对其的介绍、研究还比较少，导致研究中对共同方法偏差的控制还有所欠缺。共同方法偏差的控制方法分为程序控制和统计控制，该文介绍了共同方法偏差的多种统计控制法，如因素分析法、偏相关法、潜在的误差变量控制法、结构方程模型法等，分析了其各自的优缺点，并提出了如何选择采用这些方法的建议.
|
| [14] |
|
| [15] |
|
| [16] |
|
| [17] |
|
| [18] |
A new 4-group model of attachment styles in adulthood is proposed. Four prototypic attachment patterns are defined using combinations of a person's self-image (positive or negative) and image of others (positive or negative). In Study 1, an interview was developed to yield continuous and categorical ratings of the 4 attachment styles. Intercorrelations of the attachment ratings were consistent with the proposed model. Attachment ratings were validated by self-report measures of self-concept and interpersonal functioning. Each style was associated with a distinct profile of interpersonal problems, according to both self- and friend-reports. In Study 2, attachment styles within the family of origin and with peers were assessed independently. Results of Study 1 were replicated. The proposed model was shown to be applicable to representations of family relations; Ss' attachment styles with peers were correlated with family attachment ratings.
|
| [19] |
Two studies investigated how contextually activating attachment relationships influences the working self-concept in terms of agency and communion. In Study 1, 245 participants were primed with a secure, avoidant, or anxious-ambivalent relationship and the implicit accessibility of agency and communion was assessed using word fragments. Activating a secure relationship increased the accessibility of communion, whereas activating an anxious-ambivalent relationship increased the accessibility of agency. In Study 2, 123 participants were primed with a secure, preoccupied, avoidant-dismissive, or avoidant-fearful relationship and explicit self-perceptions of agency and communion traits were assessed using the Extended Personality Attributes Questionnaire (EPAQ). Gender interacted with the attachment prime, such that men primed with a secure relationship reported higher communion than did men primed with an avoidant (dismissive or fearful) relationship, whereas women primed with an anxious (preoccupied or fearful) relationship reported higher agency than did women primed with a secure relationship.
|
| [20] |
|
| [21] |
|
| [22] |
|
| [23] |
|
| [24] |
|
| [25] |
Despite abundant literature theorizing societal implications of algorithmic decision making, relatively little is known about the conditions that lead to the acceptance or rejection of algorithmically generated insights by individual users of decision aids. More specifically, recent findings of algorithm aversion-the reluctance of human forecasters to use superior but imperfect algorithms-raise questions about whether joint human-algorithm decision making is feasible in practice. In this paper, we systematically review the topic of algorithm aversion as it appears in 61 peer-reviewed articles between 1950 and 2018 and follow its conceptual trail across disciplines. We categorize and report on the proposed causes and solutions of algorithm aversion in five themes: expectations and expertise, decision autonomy, incentivization, cognitive compatibility, and divergent rationalities. Although each of the presented themes addresses distinct features of an algorithmic decision aid, human users of the decision aid, and/or the decision making environment, apparent interdependencies are highlighted. We conclude that resolving algorithm aversion requires an updated research program with an emphasis on theory integration. We provide a number of empirical questions that can be immediately carried forth by the behavioral decision making community.
|
| [26] |
|
| [27] |
|
| [28] |
Two studies examined the association between attachment style and perceptions of social support. Study 1 (N = 95 couples) used an experimental paradigm to manipulate social support in the context of a stressful task. Insecure participants (anxious and avoidant) who received low-support messages appraised these messages more negatively, rated a prior behavioral interaction with their partner as having been less supportive, and performed significantly worse at their task compared with secure participants. Study 2 (N = 153 couples) used a similar paradigm except that partners were allowed to send genuine support messages. Insecure participants (especially fearful) perceived their partners' messages as less supportive, even after controlling for independent ratings of the messages and relationship-specific expectations. These studies provide evidence that individuals are predisposed to appraise their support experiences in ways that are consistent with their chronic working models of attachment, especially when the support message is ambiguous.((c) 2004 APA, all rights reserved)
|
| [29] |
|
| [30] |
|
| [31] |
|
| [32] |
|
| [33] |
|
| [34] |
|
| [35] |
|
| [36] |
G*Power is a free power analysis program for a variety of statistical tests. We present extensions and improvements of the version introduced by Faul, Erdfelder, Lang, and Buchner (2007) in the domain of correlation and regression analyses. In the new version, we have added procedures to analyze the power of tests based on (1) single-sample tetrachoric correlations, (2) comparisons of dependent correlations, (3) bivariate linear regression, (4) multiple linear regression based on the random predictor model, (5) logistic regression, and (6) Poisson regression. We describe these new features and provide a brief introduction to their scope and handling.
|
| [37] |
|
| [38] |
|
| [39] |
Recommendations often play a positive role in the decision process by reducing the difficulty associated with choosing between options. However, in certain circumstances recommendations play a less positive and more undesirable role from the perspectives of both the recommending agent or agency and the person receiving the recommendation. Across a series of four studies, we explore consumer response when recommendations by experts and intelligent agents contradict the consumer's initial impressions of choice options. We find that unsolicited advice that contradicts initial impressions leads to the activation of a reactant state on the part of the decision maker. This reactance, in turn, leads to a behavioral backlash that results not only in consumers ignoring the agents' recommendations but in intentionally contradicting them.
|
| [40] |
|
| [41] |
Across 2 experiments, the authors demonstrate that emotional states influence how receptive people are to advice. The focus of these experiments is on incidental emotions, emotions triggered by a prior experience that is irrelevant to the current situation. The authors demonstrate that people who feel incidental gratitude are more trusting and more receptive to advice than are people in a neutral emotional state, and people in a neutral state are more trusting and more receptive to advice than are people who feel incidental anger. In these experiments, greater receptivity to advice increased judgment accuracy. People who felt incidental gratitude were more accurate than were people in a neutral state, and people in a neutral state were more accurate than were people who felt incidental anger. The results offer insight into how people use advice, and the authors identify conditions under which leaders, policy makers, and advisors may be particularly influential.
|
| [42] |
|
| [43] |
Attachment security is hypothesized to promote authenticity and sincerity, or honesty, whereas insecurity is hypothesized to increase various forms of inauthenticity and dishonesty. The authors tested these ideas in 8 studies of dispositional and situational attachment insecurities and their influence on inauthenticity and dishonesty. The first 4 studies showed that authenticity is related to scoring low on the 2 dimensions of dispositional attachment insecurity-anxiety and avoidance-and that these 2 dimensions are associated with different aspects of inauthenticity. The first set of studies also showed that conscious and unconscious security priming increased state authenticity (compared with neutral or insecurity priming). The last 4 studies showed that attachment insecurity is related to dishonesty (lying and cheating) and that security priming reduces the tendency to lie or cheat and does so more effectively than positive mood priming. Implications for understanding the role of authenticity and inauthenticity in various relationship contexts are discussed.2010 APA, all rights reserved
|
| [44] |
|
| [45] |
|
| [46] |
|
| [47] |
This article explores the possibility that romantic love is an attachment process--a biosocial process by which affectional bonds are formed between adult lovers, just as affectional bonds are formed earlier in life between human infants and their parents. Key components of attachment theory, developed by Bowlby, Ainsworth, and others to explain the development of affectional bonds in infancy, were translated into terms appropriate to adult romantic love. The translation centered on the three major styles of attachment in infancy--secure, avoidant, and anxious/ambivalent--and on the notion that continuity of relationship style is due in part to mental models (Bowlby's "inner working models") of self and social life. These models, and hence a person's attachment style, are seen as determined in part by childhood relationships with parents. Two questionnaire studies indicated that relative prevalence of the three attachment styles is roughly the same in adulthood as in infancy, the three kinds of adults differ predictably in the way they experience romantic love, and attachment style is related in theoretically meaningful ways to mental models of self and social relationships and to relationship experiences with parents. Implications for theories of romantic love are discussed, as are measurement problems and other issues related to future tests of the attachment perspective.
|
| [48] |
|
| [49] |
|
| [50] |
Models can convey biases and false information to users.
|
| [51] |
|
| [52] |
Even though computational algorithms often outperform human judgment, received wisdom suggests that people may be skeptical of relying on them (Dawes, 1979). Counter to this notion, results from six experiments show that lay people adhere more to advice when they think it comes from an algorithm than from a person. People showed this effect, what we call algorithm appreciation, when making numeric estimates about a visual stimulus (Experiment 1A) and forecasts about the popularity of songs and romantic attraction (Experiments 1B and 1C). Yet, researchers predicted the opposite result (Experiment 1D). Algorithm appreciation persisted when advice appeared jointly or separately (Experiment 2). However, algorithm appreciation waned when: people chose between an algorithm's estimate and their own (versus an external advisor's; Experiment 3) and they had expertise in forecasting (Experiment 4). Paradoxically, experienced professionals, who make forecasts on a regular basis, relied less on algorithmic advice than lay people did, which hurt their accuracy. These results shed light on the important question of when people rely on algorithmic advice over advice from people and have implications for the use of "big data" and algorithmic advice it generates.
|
| [53] |
Rapid development and adoption of AI, machine learning, and natural language processing applications challenge managers and policy makers to harness these transformative technologies. In this context, the authors provide evidence of a novel “word-of-machine” effect, the phenomenon by which utilitarian/hedonic attribute trade-offs determine preference for, or resistance to, AI-based recommendations compared with traditional word of mouth, or human-based recommendations. The word-of-machine effect stems from a lay belief that AI recommenders are more competent than human recommenders in the utilitarian realm and less competent than human recommenders in the hedonic realm. As a consequence, importance or salience of utilitarian attributes determine preference for AI recommenders over human ones, and importance or salience of hedonic attributes determine resistance to AI recommenders over human ones (Studies 1–4). The word-of machine effect is robust to attribute complexity, number of options considered, and transaction costs. The word-of-machine effect reverses for utilitarian goals if a recommendation needs matching to a person’s unique preferences (Study 5) and is eliminated in the case of human–AI hybrid decision making (i.e., augmented rather than artificial intelligence; Study 6). An intervention based on the consider-the-opposite protocol attenuates the word-of-machine effect (Studies 7a–b).
|
| [54] |
|
| [55] |
|
| [56] |
|
| [57] |
Personalized interventions are deemed vital given the intricate characteristics, advancement, inherent genetic composition, and diversity of cardiovascular diseases (CVDs). The appropriate utilization of artificial intelligence (AI) and machine learning (ML) methodologies can yield novel understandings of CVDs, enabling improved personalized treatments through predictive analysis and deep phenotyping. In this study, we proposed and employed a novel approach combining traditional statistics and a nexus of cutting-edge AI/ML techniques to identify significant biomarkers for our predictive engine by analyzing the complete transcriptome of CVD patients. After robust gene expression data pre-processing, we utilized three statistical tests (Pearson correlation, Chi-square test, and ANOVA) to assess the differences in transcriptomic expression and clinical characteristics between healthy individuals and CVD patients. Next, the recursive feature elimination classifier assigned rankings to transcriptomic features based on their relation to the case–control variable. The top ten percent of commonly observed significant biomarkers were evaluated using four unique ML classifiers (Random Forest, Support Vector Machine, Xtreme Gradient Boosting Decision Trees, and k-Nearest Neighbors). After optimizing hyperparameters, the ensembled models, which were implemented using a soft voting classifier, accurately differentiated between patients and healthy individuals. We have uncovered 18 transcriptomic biomarkers that are highly significant in the CVD population that were used to predict disease with up to 96% accuracy. Additionally, we cross-validated our results with clinical records collected from patients in our cohort. The identified biomarkers served as potential indicators for early detection of CVDs. With its successful implementation, our newly developed predictive engine provides a valuable framework for identifying patients with CVDs based on their biomarker profiles.
|
| [58] |
|
| [59] |
Five studies examined the effects of priming the secure base schema on intergroup bias. In addition, Studies 1-2 examined the effects of dispositional attachment style, Studies 2-5 examined a mood interpretation, Study 3 examined the mediating role of threat appraisal, and Studies 4-5 examined the effects of secure base priming while inducing a threat to self-esteem or cultural worldview. Secure base priming led to less negative evaluative reactions toward out-groups than positive affect and neutral control conditions. In addition, whereas the effects of secure base priming did not depend on attachment style and were not explained by mood induction, they were mediated by threat appraisal and occurred even when self-esteem or cultural worldview was threatened. The discussion emphasizes the relevance of attachment theory for understanding intergroup attitudes.
|
| [60] |
|
| [61] |
|
| [62] |
|
| [63] |
|
| [64] |
Whereas attachment dimensions (i.e., anxiety and avoidance) are generally associated with lower levels of relationship evaluations (e.g., satisfaction, commitment), research has not yet fully incorporated how partner attachment is related to these evaluations, nor how dyadic patterns (actor × partner attachment interactions) are associated with evaluations. Across two dyadic studies ( = 185, 123 dyads), we examine how actor, partner, and actor × partner interactions of attachment anxiety and avoidance are associated with reports of trust, satisfaction, and commitment. Results generally revealed that actor effects of attachment anxiety on lower relationship evaluations were weaker when partners were more anxious and stronger when partners were more avoidant. Moreover, actor effects of attachment avoidance on lower trust and satisfaction were stronger when partners were more anxious. Finally, own avoidance was more strongly negatively related to commitment in the presence of a more avoidant partner. These results suggest that the combination of attachment within relationships is important to consider for both close relationships researchers and clinicians.
|
| [65] |
|
| [66] |
|
| [67] |
Objective. To better understand 1) why patients have a negative perception of the use of computerized clinical decision support systems (CDSSs) and 2) what contributes to the documented heterogeneity in the evaluations of physicians who use a CDSS. Methods. Three vignette-based studies examined whether negative perceptions stemmed directly from the use of a computerized decision aid or the need to seek external advice more broadly (experiment 1) and investigated the contributing role of 2 individual difference measures, attitudes toward statistics (ATS; experiment 2) and the Multidimensional Health Locus of Control Scale (MHLC; experiment 3), to these findings. Results. A physician described as making an unaided diagnosis was rated significantly more positively on a number of attributes than a physician using a computerized decision aid but not a physician who sought the advice of an expert colleague (experiment 1). ATS were unrelated to perceptions of decision aid use (experiment 2); however, greater internal locus of control was associated with more positive feelings about unaided care and more negative feelings about care when a decision aid was used (experiment 3). Conclusion. Negative perceptions of computerized decision aid use may not be a product of the need to seek external advice more generally but may instead be specific to the use of a nonhuman tool and may be associated with individual differences in locus of control. Together, these 3 studies may be used to guide education efforts for patients.
|
| [68] |
|
| [69] |
|
| [70] |
|
| [71] |
|
| [72] |
The relationship between trust, confidence, and expertise in Judge-Advisor Systems is examined in two experiments with Judge-Advisor pairs, one with strangers and another with participants in ongoing relationships. There was expertise asymmetry so that Judges had less expertise than their Advisors. The dyads could receive money for accurate Judge decisions. Either the Judge or Advisor had the power to allocate this money between dyad members, before task interaction in study one and after task completion in study two. Because Judges were more dependent on Advisors than vice versa, it was predicted that trust would be more important to Judges. Results were supportive. Judges had higher and more variable ratings of trust in their partner than did Advisors, suggesting that Judges were more motivated to evaluate trust. High confidence by Advisors had a positive impact on Judges' ratings of trust and tendency to follow their advice. Judges' trust in their Advisors was significantly related their taking the advice and being confident in their final decisions. Although participants in study two had higher levels of trust in their partners, they allocated less money to them. The implications for establishing trust are discussed. Copyright 2001 Academic Press.
|
| [73] |
|
| [74] |
From fake social media accounts and generative artificial intelligence chatbots to trading algorithms and self-driving vehicles, robots, bots and algorithms are proliferating and permeating our communication channels, social interactions, economic transactions and transportation arteries. Networks of multiple interdependent and interacting humans and intelligent machines constitute complex social systems for which the collective outcomes cannot be deduced from either human or machine behaviour alone. Under this paradigm, we review recent research and identify general dynamics and patterns in situations of competition, coordination, cooperation, contagion and collective decision-making, with context-rich examples from high-frequency trading markets, a social media platform, an open collaboration community and a discussion forum. To ensure more robust and resilient human-machine communities, we require a new sociology of humans and machines. Researchers should study these communities using complex system methods; engineers should explicitly design artificial intelligence for human-machine and machine-machine interactions; and regulators should govern the ecological diversity and social co-development of humans and machines.© 2024. Springer Nature Limited.
|
| [75] |
|
| [76] |
I explored advice acceptance for high‐stakes decisions (i.e., those with subjectively important and risky outcomes), focusing on the relative influence of two components of consumer trust—benevolence and expertise—as well as perceived emotional decision difficulty. Participants solicited advice from experts when their decisions were low in perceived emotional difficulty but favored the advice of predominantly benevolent providers when making highly emotionally difficult decisions. Although consumers who faced emotionally difficult decisions were willing to trade off expertise for benevolence, they did not perceive this non‐normative trade‐off to influence decision quality. Instead, the results support a “stress buffering” effect whereby consumers were more confident in the accuracy of predominantly benevolent providers’ advice.
|
| [77] |
|
| [78] |
Generative artificial intelligence (GenAI) holds the potential to transform the delivery, cultivation and evaluation of human learning. Here the authors examine the integration of GenAI as a tool for human learning, addressing its promises and challenges from a holistic viewpoint that integrates insights from learning sciences, educational technology and human-computer interaction. GenAI promises to enhance learning experiences by scaling personalized support, diversifying learning materials, enabling timely feedback and innovating assessment methods. However, it also presents critical issues such as model imperfections, ethical dilemmas and the disruption of traditional assessments. Thus, cultivating AI literacy and adaptive skills is imperative for facilitating informed engagement with GenAI technologies. Rigorous research across learning contexts is essential to evaluate GenAI's effect on human cognition, metacognition and creativity. Humanity must learn with and about GenAI, ensuring that it becomes a powerful ally in the pursuit of knowledge and innovation, rather than a crutch that undermines our intellectual abilities.© 2024. Springer Nature Limited.
|
| [79] |
|
| [80] |
Our framework for understanding advice-taking in decision making rests on two theoretical concepts that motivate the studies and serve to explain the findings. The first is egocentric discounting of others' opinions and the second is reputation formation for advisors. Advice discounting is attributed to differential information, namely, the notion that decision makers have privileged access to their internal reasons for holding their own opinion, but not to the advisors' internal reasons. Reputation formation is related to the negativity effect in impression formation and to the trust asymmetry principle. In three studies we measured decision makers' weighting policy for advice and, in a fourth study, their willingness to pay for it. Briefly, we found that advice is discounted relative to one's own opinion, while advisors' reputations are rapidly formed and asymmetrically revised. The asymmetry implies that it may be easier for advisors to lose a good reputation than to gain one. The cognitive and social origins of these phenomena are considered. Copyright 2000 Academic Press.
|
| [81] |
|
| [82] |
|
/
| 〈 |
|
〉 |
