Artificial intelligence (AI) systems increasingly serve as advisors, evaluators, and decision-makers, yet little is known about how people perceive AI as a social entity. Across five primary studies and eight supplementary studies, we show that people assign racial identities to AI systems, overwhelmingly perceiving them as White-even though these systems provide no visual, vocal, or identity cues from which race might ordinarily be inferred. Using forced-choice, open-ended, and implicit reverse-correlation measures, Studies 1 and 2 demonstrate that AI is explicitly and implicitly associated with Whiteness across diverse samples. Study 3 extends these findings beyond the United States, showing that AI is perceived as White in Japan and India. Study 4 examines the implications of AI racialization, showing that perceiving an AI system as more White is associated with greater trust in and persuasiveness of its recommendations. Supplementary studies identified two potential mechanisms: stereotype spillover linking intelligence with Whiteness and ecosystem-based inferences based on beliefs about who creates, trains, and uses AI. Building on this account, Study 5 provides causal evidence by isolating a key ecosystem cue-the racial composition of AI training data. Participants assigned less cognitively demanding tasks to AI systems described as trained on Black and Latino data than to otherwise identical systems trained on White or unspecified data. Together, these findings show that AI is not perceived as socially neutral but instead acquires racial meanings associated with credibility, authority, and capability, demonstrating how social categories shape perceptions of novel technological entities beyond their underlying algorithmic properties. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
The development of narcissism has been theoretically tied to childhood experiences, yet little empirical research has examined the influence of the family environment on adult narcissism using long-term longitudinal data. The present study identified family factors in late childhood and early adolescence that are prospectively associated with narcissism (grandiosity, superiority, exploitativeness, vulnerability) in emerging adulthood using data from a longitudinal study of Mexican-origin youth (N = 674) spanning from age 10 to 26. Using structural equation modeling of multimethod data (child report, parent and spouse report, and coded parent-child interactions), we found that grandiosity and superiority were significantly predicted by positive aspects of the family environment (e.g., parental warmth, parent-child relationship quality, parental relationship quality), whereas exploitativeness and vulnerability were significantly predicted by negative aspects of the family environment (e.g., parental hostility, parent-child conflict). These findings were much weaker and often nonsignificant after controlling for earlier levels of narcissism, suggesting that the family environment exerts its effect on adult narcissism via its concurrent effect on adolescent narcissism. Overall, our study highlights several underexplored family factors in childhood that contribute to the development and maintenance of narcissism in adulthood. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
People frequently encounter minor aversive experiences-such as rude interactions or social slights-that, while not major events, can still affect psychological well-being. In this research, we examine how people experience and perceive social support from artificial intelligence (AI) following such experiences. Drawing on mind perception theory, we propose that initial expectations about AI's experiential capacity relative to humans shape how people experience and perceive AI support. A pilot study demonstrates that people have different initial expectations of AI's versus humans' experiential capacity. Across six studies (five preregistered; total N = 3,341), we found that these initial expectations produce two distinct effects: People report lower fear of negative evaluation after receiving support from AI versus a human, which in turn improves well-being outcomes (Studies 1-3), and people perceive AI as exceeding their expectations for empathy, which in turn drives perceived support effectiveness (Studies 4-6). These effects emerged across controlled simulations and real interactions with actual human and AI support givers. Together, these findings reveal that initial psychological expectations about AI's experiential capacity play a central role in shaping the experience and perception of AI support. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
This preregistered meta-analysis of longitudinal studies investigated the prospective effects between self-esteem and aggression. The link between these constructs has been debated for decades. Some researchers have argued that low self-esteem is a risk factor for aggression, while others have linked high self-esteem to aggression, referring to a "dark side of high self-esteem." A comprehensive literature search led to the inclusion of 47 samples (total N = 40,688). Using one-stage meta-analytic structural equation modeling, we estimated dynamic panel models controlling for stable between-person variance in the constructs. The analyses revealed no significant cross-lagged effects; self-esteem did not predict later aggression, nor did aggression predict later self-esteem. When using continuous time meta-analysis to account for between-study differences in time intervals, a significant but small effect of aggression on low self-esteem emerged. The effect of self-esteem on aggression remained nonsignificant. In both types of analyses, the latent factors of self-esteem and aggression were significantly correlated with a large effect size (r = -.35 and r = -.33, respectively). There was little evidence for publication bias. The only evidence of moderation of cross-lagged effects was found for the effect of aggression on self-esteem, which was stronger (i.e., more negative) for hostile-only aggression. Otherwise, the results provided no clear evidence that the prospective effects between self-esteem and aggression differed across aggression dimensions, age, and gender. Taken together, our results indicated that neither low nor high self-esteem appears to be a risk factor for future aggression. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
People tend to morally prioritize their extremely close others (loved ones; such as family, partners, or friends) while disregarding extremely distant others (societal outcasts; such as pariahs, villains, or criminals). Across seven primary and 15 supplemental studies, we document how multicultural experiences might alter this universal default through moral compression: the simultaneous expansion of moral concern toward societal outcasts and contraction away from loved ones, effectively "compressing" the outermost and innermost anchors of the moral circle closer toward each other. Specifically, we find that greater breadth of living abroad experiences can compress the moral circle through heightened identification with all humanity, an association strengthened by cultural distance between home and host countries. Moral compression reflects a shift toward more impartial moral treatment under cognitive constraints. Whereas centuries of moral thought have idealized an ever-expanding moral circle, we draw from theories of bounded cognition and social identity abstraction to explore a costly trade-off of moral change. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Across nine experiments featuring both hypothetical donations and incentive-compatible designs, we examine how benefactor income shapes observers' emotional and behavioral responses to witnessing prosocial behavior. We show that identical acts of giving elicit different levels of moral elevation and downstream generosity depending on who gives. Specifically, generosity from low-income (vs. high-income) benefactors is perceived as more altruistically motivated, which elicits stronger feelings of moral elevation and, in turn, prosocial contagion. We theorize that these effects occur not because high-income benefactors are demotivating but because low-income benefactors reduce observers' skepticism, consistent with a virtue discounting account. These effects are context dependent such that the presence of high-income benefactors attenuates moral elevation and generosity even when low-income benefactors are also present, whereas making altruistic motives salient attenuates income-based differences in moral elevation and subsequent giving. We further show that these effects are not fully explained by alternative mechanisms such as general positive affect, guilt, or self-reflection. By linking income-based motive inferences, virtue discounting, and moral elevation, this research advances understanding of how prosocial behavior spreads socially and why identical acts of generosity can inspire markedly different downstream responses. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Declining institutional trust across democracies raises pressing questions about its social consequences. We examine how institutional distrust shapes leadership preferences. We propose that institutional distrust creates a dissonant state. When institutions are seen as violating normative expectations of benevolence, individuals experience cognitive conflict between distrusting institutions and believing that they ought to serve the public interest. We argue that such a conflict heightens the need for cognitive consistency, which in turn increases preferences for leaders who express one-sided, unambivalent stances. We test this account across five sets of studies conducted in eight countries. Studies 1A and 1B establish that institutional distrust elicits cognitive conflict and that this effect is stronger among individuals who strongly believe that institutions should act in the best interest of all. Studies 2A, 2B, and 2C demonstrate that cognitive conflict mediates the effect of institutional distrust on the need for cognitive consistency using experimental designs and cross-national samples, with Study 2C further examining self-relevance as a moderator. Studies 3A and 3B show that heightened need for cognitive consistency predicts stronger preferences for one-sided leaders. Studies 4A and 4B provide evidence for the full serial mediation model across experimental and correlational designs. Finally, Study 5 uses reverse correlation to demonstrate that under institutional distrust, the mental representation of a prototypical leader appears more one-sided. Together, these findings identify cognitive consistency as a key mechanism through which rising institutional distrust reshapes leadership preferences. More broadly, they support cognitive dissonance theory by identifying its process in one topical context. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
People often anticipate feared, negative future events but still fail to take preventive action-simply imagining feared outcomes does not reliably prompt behavior change. Research has shown that mental contrasting-imagining a negative future and then reflecting on a positive reality in the present-can translate this fear into action. However, the psychological mechanisms underlying this effect remain unexplored. Across five experiments (two cross-sectional and three longitudinal with 1-month follow-ups), we examined how mental contrasting of a negative future promotes preventive behavior by altering interpretations of current reality. Supporting our hypothesis, participants who engaged in mental contrasting evaluated their reality as more precious than those engaged in reverse contrasting (Studies 1 and 3) and those in the valence control condition (Study 2). These shifts in evaluations of reality predicted greater preventive behavior over the following month (Studies 1 and 3). In addition, mental contrasting changed implicit evaluations of reality in the same direction, as measured by the extrinsic affective Simon task (Studies 4 and 5). These implicit changes further predicted preventive action 1 month later (Study 5). Our research identifies a key mechanism by which mental contrasting converts fear into action: It changes how people explicitly and implicitly evaluate the meaning of their current reality. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
The reciprocal influence between personality and relationship experiences has been of considerable interest, but there is little research on the extent to which changes in relationship quality are shaped by changes in antagonistic or self-centered personality traits or, conversely, how changes in relationship quality impact these traits. In the present study, we examined how Machiavellianism, psychopathy, and narcissism are linked to key aspects of romantic relationships, including relationship satisfaction, commitment, reciprocity orientation, and verbal aggression. We used data from the German Family Panel (pairfam). The analyzed sample included N = 1,925 couples, with a mean age of 36.48 years (SD = 7.62) for anchor persons and 36.91 years (SD = 8.71) for their partners at the first measurement occasion. Across three waves of data collection, each 2 years apart, we applied extended bivariate latent change models. The results revealed both cross-sectional associations and correlated changes between the antagonistic/self-centered traits and relationship satisfaction, commitment, reciprocity orientation, and verbal aggression, with both actor and partner effects. Generally, higher levels in these traits were linked to lower relationship satisfaction and commitment, alongside higher reciprocity orientation and verbal aggression. Moreover, findings indicated a reciprocal influence between personality (change) and relationship (change), again involving both actor and partner effects, with narcissism showing the most consistent associations with relationship dynamics. The results are discussed in the context of existing literature on personality development in partner relationships. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
In first social encounters, people regularly try to figure out whether others like them. When accurate, these meta-liking judgments enable individuals to more optimally engage in social interactions. Although previous research shows that people do know how much others like them, little is known about the processes leading to accurate meta-liking judgments. In the present, pre-registered research, we aimed to close this gap by investigating the role of a meta-perceiver’s appearance and behavior, others’ behavior towards that meta-perceiver, and the reciprocity heuristic in forming accurate meta-liking judgments. To this end, we used data from a round-robin study, in which N = 144 participants (Mage = 22.91, 67.6% female) took part in two group meetings, where they interacted with unacquainted others. After each participant had given a short self-introduction, participants indicated liking and meta-liking for each other participant. Additionally, trained raters judged participants’ appearance and behavior during the self-introduction. Results showed that people were able to correctly infer how much others generally liked them but hardly knew how much specific other liked them. Lens model analyses revealed that although one’s appearance and behavior were valid cues for being generally liked, meta-perceivers did not utilize most of these observable cues to inform their meta-liking judgments. However, the assumed reciprocity mediated both individual and dyadic meta-liking accuracy. In light of these findings, we discuss the role of self-observed behavior and stable personality characteristics as well as situational and motivational factors, presenting an integrative theoretical framework to explain generalized and dyadic meta-liking accuracy.
Prior research examining the impact of job loss on personality change has yielded inconsistent findings. The present preregistered study extends prior research by analyzing longitudinal data from Sample N = 1,110 predominantly first-generation Mexican-origin adults, who reported on their employment status and Big Five personality traits up to eight times over 16 years. We explored the influence of both individual and spousal unemployment on four forms of personality trait stability and change: rank-order stability, mean-level changes, within-person trait stability, and profile stability. Job loss was associated with a wide range of mean-level personality changes; collectively, the findings suggest that the experience of unemployment can disrupt the normative process of maturation, leading to decreases in conscientiousness and agreeableness and impeding the typical decline in neuroticism. Our study is the first to establish a spillover effect of spousal job loss on partners' personality dynamics, demonstrating that personality changes can occur in response to a partner's job loss. In some cases, these changes paralleled those found for one's own job loss (specifically, gradual postevent declines in agreeableness and openness), whereas in other cases, we found evidence of a compensatory effect (where individuals whose partner lost their job showed an immediate increase in conscientiousness). On the contrary, job loss had no significant effects on rank-order stability, whether experienced by self or spousal, and was associated with higher extraversion within-person trait stability. Likewise, job loss was associated with lower profile stability; those who experienced unemployment showed lower stability in their personality profiles than those who remained employed. These findings highlight the importance of studying the effects of major life transitions on personality change in diverse samples. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Having outgroup friends not only improves attitudes toward the (primary) outgroup but potentially also toward other (secondary) outgroups. This "secondary transfer effect" is attributed to the generalization of intergroup contact, but it may also reflect social influence as friends shape each other's outgroup attitudes. In three studies (two empirical and one simulation), we applied stochastic actor-oriented models to longitudinal network data representing friendships between adolescents. Across Studies 1 (n = 812) and 2 (n = 2,559), we found no evidence that having more primary outgroup friends was associated with a change in secondary outgroup attitudes; instead, adolescents' attitudes became more similar over time to those of their primary outgroup friends. Study 3, a simulation, further showed that social influence can lead to either more positive or more negative attitudes toward secondary outgroups. Thus, friendships with primary outgroup members may have negative implications for secondary outgroups when those friends hold more negative attitudes. We discuss how social influence can advance research on the generalization of intergroup contact effects. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Advances in generative artificial intelligence (AI) have given rise to relational AI-AI agents that mimic human relational capabilities while possessing unique nonhuman features, including constant availability, extensive malleability, and a lack of intrinsic psychological needs. This article uses self-determination theory to examine how relational AI may support or undermine the three basic psychological needs self-determination theory posits as essential for well-being: relatedness, competence, and autonomy. Across roles including that of tutor, copilot, social mediator, companion, and therapist, relational AI may address critical challenges to these needs, facilitating goal attainment, alleviating loneliness, and promoting mental health. However, relational AI also carries potential risks for well-being, including reduced self-direction, diminished efficacy, and altered expectations within human relationships. We propose that relational AI's ultimate impact on users' well-being will depend on moderating factors that may shape both the strength and direction of its effects on well-being, including users' motivational orientation toward the goal for which they use relational AI, their motivational orientation for using relational AI, their baseline psychological need satisfaction, dynamics particular to each relational AI role, and various situational factors. Crucially, it is likely impossible to fully understand relational AI's impacts on well-being without examining all three psychological needs, as focusing on only a single need risks overlooking dynamics in which impacts on multiple needs interact to amplify or neutralize well-being. Future research on these processes can deepen our understanding of human relationship dynamics, inform responsible AI development, and reveal novel theoretical mechanisms underlying relational AI's impacts on well-being. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
People regularly disagree, but disagreements come at a cost to relationships and well-being when communication is poor. We posited that high-quality listeners who convey their clear attention, understanding, and positive intention can turn a disagreement into a constructive experience for speakers and those expressing their views. Informed by self-determination theory, we examined whether receiving high-quality listening would improve speakers' state well-being (SWB) and its future-oriented component, state optimism, by satisfying the psychological needs for relatedness, autonomy, and competence. Across five studies (N = 1,624), three of which were preregistered, and using diverse methods including experimental vignettes, live conversations, recalled conversations, and a multilevel field survey, we compared high-quality, moderate-quality (Studies 1-5), and low-quality (Study 1) listening when conversing partners agreed or disagreed. Speakers who experienced high-quality listening consistently reported greater SWB, state optimism, and basic psychological need satisfaction. The main effects on SWB and state optimism were consistently mediated by relatedness need satisfaction. Although the indirect effects via autonomy and competence varied, an internal meta-analysis showed that both significantly mediated SWB and state optimism. Importantly, in Studies 3 and 4, high-quality listening during a disagreement resulted in higher speakers' SWB than moderate-quality listening during an agreement. These findings challenge the assumption that disagreement is inherently divisive and extend interpersonal relationship research by providing a better understanding of its consequences. They point to listening quality as a pathway to SWB and social connection, even when views are opposing. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Loneliness and social isolation are associated with numerous adverse physical and psychological health outcomes in older adulthood, including cognitive impairment and mortality risk. Yet, how the individual and joint effects of loneliness and social isolation contribute to these outcomes remains unclear, particularly given the interplay between individual differences (loneliness) and environmental factors (social isolation) in shaping these important health outcomes in older adulthood (i.e., person-environment interactions). We used Cox regression, logistic regression, and multistate survival models to systematically investigate the individual and adjusted associations among loneliness, social isolation, cognitive aging outcomes, and mortality risk. Further, extensive between-study variability in operational definitions, modeling approaches, and covariate adjustments may be contributing to mixed results in the existing literature. In this registered report, we applied a multistudy approach (i.e., coordinated data analysis) in which independent but identical models were fit across 11 longitudinal studies representing participants from 18 countries (Ntotal = 175,070). Random effects meta-analyses synthesized results across studies, showing that loneliness was consistently associated with an elevated risk of cognitive impairment and mortality across statistical approaches, even after adjusting for social isolation. Conversely, social isolation was not consistently associated with cognitive impairment and showed weaker associations with mortality risk. Together, our findings suggest that loneliness is a robust, proximal predictor of major aging outcomes and highlight avenues for theory development, strategies to strengthen cognitive resilience and independence in older adulthood, and more efficient resource allocation of public health resources. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Moralized attitudes motivate social change, political participation, and charitable giving, but at the same time, hinder compromise, conformity, and acceptance of procedural solutions to conflict. Although considerable research has explored the consequences of moralized attitudes, the processes that lead to attitude moralization and demoralization are less clear. Across six studies (N = 4,464), we contrasted the moralization of attitudes about male circumcision (typically viewed as a matter of preference or convention in the United States) with the demoralization of attitudes about female circumcision (a practice most Americans see as immoral). For male circumcision, stronger negative emotional reactions consistently mediated the association between exposure to evocative information and stronger moral conviction, even when accounting for participants' underlying levels of support or opposition. For female circumcision, demoralization was observed but less consistently across studies, and mediational evidence emerged only under stronger informational "pushes": No individual piece of information was potent enough to predict demoralization on its own when participants chose which information to read, but weaker negative emotional reactions fully mediated the effect of information exposure on reduced moral conviction when participants were randomly assigned to read all (vs. no) information. These findings suggest that although negative emotion plays a central role in both moralization and demoralization, demoralization is harder to achieve and may require more extensive information exposure. In other words, attitudes may acquire their moral significance more readily than they lose it. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Extensive evidence suggests that war-related trauma negatively affects health, yet its long-term and transgenerational effects on psychological and physical adjustment remain poorly understood. This study examines whether individuals who experienced greater war exposure in early childhood-specifically variation in bombardment intensity across German municipalities during the Second World War-show lower, similar, or better psychological (life satisfaction, affective well-being, and mental health) and physical adjustment (self-rated and physical health) in late adulthood compared to less exposed peers. The study also analyzes intergenerational effects by investigating whether the children of parents who were more versus less exposed to bombardments differed in psychological and physical adjustment in midlife. War exposure was measured using exogenous variation in regional housing destruction across 1,739 German municipalities (0%-97% destroyed) during the Second World War. Combining these historical data with individual data from the German Socio-Economic Panel, we analyzed three samples: two childhood-exposed samples (N = 1,975, Mage = 73.44; N = 959, Mage = 72.09) and one consisting of their children (N = 1,747, Mage = 51.67). Supporting perspectives on individual resilience, the findings revealed no significant associations between early-life war exposure and late-life adjustment. Likewise, parents' war experiences were not associated with their children's midlife adjustment. Process analyses, including socioeconomic status, control beliefs, and general stress responses only partially accounted for the observed associations. These results challenge the notion of inevitable long-term or intergenerational harm from early-life trauma and highlight the need for future replication studies examining individual differences in responses to war exposure. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
When thinking about a task or goal feels difficult, people can interpret this as signaling self-relevant value (difficulty-as-importance) and a reminder to stop wasting time on self-irrelevant things (difficulty-as-impossibility). Identity-based motivation theory predicts that both interpretations are available in memory, vary in momentary accessibility, and shape action in part through activating action-relevant judgments. We note that in situations calling for task engagement, action-relevant judgment entails making sense of time, an abstract concept that people reason about metaphorically. We test this novel prediction that difficulty mindsets shape action in part by shaping how people reason metaphorically about time by examining associations (N = 941) and documenting effects of mindset accessibility (using an autobiographical recall task, N = 1,676) on how much people endorse the time-as-limited and time-as-expandable aspects of the time-as-resource metaphor. Indeed, across experiments, people more strongly endorsed time-as-limited, less strongly endorsed time-as-expandable and saw these aspects of the time-as-resource metaphor as more negatively correlated if randomized to the difficulty-as-impossibility (vs. difficulty-as-importance) recall condition. Supporting our novel prediction, the consequence of accessible difficulty mindsets on task engagement (Experiments 2-3) and performance (Experiments 4-5), confidence in goal progress and experienced time to make progress (the length of the line they drew to represent the semester or year, Experiments 6-7) occurred partly through the effect of accessible mindsets on momentary judgments of time-as-limited and time-as-expandable, as revealed in statistical mediation analyses. In situations requiring task engagement and assessing goal certainty, accessible difficulty mindsets matter in part by momentarily shifting judgments about having time. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Research suggests that people who have relatively secure attachment styles tend to experience better mental and physical health across the lifespan (Pietromonaco & Beck, 2019; X. Zhang et al., 2022). Despite the growing body of research linking attachment styles to health outcomes, relatively little is known about why or how these associations arise. The present study analyzed data from 1,377 older adult twins, who completed measures of their general and relationship-specific attachment styles and five health assessments-depressive symptoms, self-reported mental and physical health, physical illnesses, and waist circumference. Leveraging the study's twin design, we estimated the relative contributions of genetic and environmental factors to the associations between attachment styles and health. Our findings suggest that the genetic factors underlying depression are largely overlapping with the genetic factors underlying attachment anxiety (e.g., genetic correlation, ra, = .70 between general attachment anxiety and depression) and show a moderate degree of overlap with the genetic factors underlying attachment avoidance (average ra = .34). In general, genetic and environmental factors each accounted for approximately half of the total phenotypic associations between attachment styles and mental health outcomes. Furthermore, we found moderate genetic correlations between attachment anxiety and physical health outcomes (average |ra| = .30), but little-to-no overlap among the environmental factors that contribute to attachment anxiety and physical health. Thus, the associations between attachment anxiety and physical health outcomes appear to be primarily attributable to a common or shared genetic liability. Attachment avoidance shared weak or near-zero associations with indicators of physical health. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
