About

Molly Crockett is a Professor in the Department of Psychology at Princeton University. She holds a Ph.D. from the University of Cambridge. Her research investigates the relationships between self and society, power and knowledge, technology and culture. Her lab seeks to understand the nature of epistemic injustice, exploring how systems of power influence our understanding of the world and how knowledge is marginalized or amplified through social learning, technology, and their interactions. Crockett's work particularly focuses on how technologies marketed as artificial intelligence reinforce existing social inequalities and power structures. Her research integrates theoretical perspectives from cognitive science, social psychology, philosophy, science and technology studies, and data science. She employs a variety of methods, including behavioral experiments in laboratory and online settings, field studies, computational modeling, and machine learning. Her goal is to bridge quantitative and qualitative approaches to better understand human experience and societal transformation.

Research topics

• Political Science
• Psychology
• Social Science
• Sociology
• Social psychology
• Medicine
• Public relations
• Engineering
• Virology
• Psychiatry
• Computer Science
• Artificial Intelligence
• Cognitive science
• Physics
• Cognitive psychology
• Psychoanalysis
• Data science
• Internal medicine
• Neuroscience
• Demography
• Clinical psychology

Selected publications

• Hypocritical blame is associated with reduced prosocial motivation

  Scientific Reports · 2025-09-25

  articleOpen access

  People often act hypocritically. One form of hypocrisy occurs when people blame others for transgressing moral principles they themselves have violated in the past. However, the psychological processes linked to this hypocritical blame are largely unknown. One possibility is that hypocritical blame is associated with the costs of being prosocial, such that a person could intend to help but is unwilling to put in the effort. Here, we test whether a measure of hypocritical blame that quantifies the discrepancy between willingness to profit from another's harm, and blaming somebody else for similarly profiting, is related to the motivation to choose and then exert physical effort to benefit themselves or a stranger. Results revealed that hypocritical blame is associated with reduced prosocial motivation specifically, and not with how willing people are to exert effort for their own benefit. This effect was found in both a reduced willingness to choose to be prosocial and for energising prosocial acts. This suggests that the discrepancy between moral standards and actions is related to the willingness to overcome the costs of being prosocial, with some people being simply unwilling to exert the effort required to live up to their moral principles.

  PublisherOA PDFDOI

• Instrumental harm and impartial beneficence distinctively frame cognitive representations of moral decision problems.

  Journal of Experimental Psychology General · 2025-11-01

  articleOpen accessSenior author

  Utilitarian ethical theories argue that the morality of actions depends on their consequences for impartially maximizing overall welfare. Recent research suggests that individual differences in utilitarian tendencies fall along two dimensions: a permissive attitude toward harming others for greater good (instrumental harm [IH]) and an impartial concern for others' welfare (impartial beneficence [IB]). We hypothesize that these dimensions operate as intuitive theories in the moral domain, framing distinctive patterns of moral judgments and behavior. Using intersubject representational similarity analysis of behavioral data (N = 254), we found that when participants shared endorsement of instrumental harm or impartial beneficence, they showed similar patterns of moral judgment and decision making. Intersubject representational similarity analysis of functional neuroimaging data (N = 68) revealed that participants with similar endorsement of instrumental harm or impartial beneficence showed similar neural encoding of moral choice attributes, even when they made different choices. Meanwhile, participants with dissimilar endorsement of these dimensions showed distinctive neural encoding of moral choice attributes, even when they made similar choices. These similarity and dissimilarity patterns emerged in distinct brain regions for instrumental harm and impartial beneficence. Together, our findings suggest that instrumental harm and impartial beneficence distinctively frame cognitive representations of moral decision problems, over and above guiding judgments and decisions. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

  PublisherDOI

• Introspective access to value-based multi-attribute choice processes

  Nature Communications · 2025-04-20 · 2 citations

  articleOpen accessSenior author

  People routinely choose between options varying on multiple attributes - homes to rent, movies to watch, and so on. Here, we test how much awareness people have of the mental processes underlying these choices. We develop a method to quantify awareness of value-based multi-attribute choice processes that accounts for diverse choice strategies. Across five studies, participants make choices and then report how they believe they made them. We use computational modeling to identify the process revealed in their choices, and compare it to their self-reports to quantify individuals' accuracy about their choice process. While we observe substantial variation in accuracy, participants are often highly accurate about their choice process - more accurate than predicted by a sample of decision scientists - and more accurate than informed third-party observers, suggesting evidence for introspection. These results challenge notions that we are strangers to ourselves and instead suggest that people often know how they made value-based choices.

  PublisherOA PDFDOI

• Neural signatures of harm aversion predict later willingness to exert effort for others’ rewards

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-10-04

  preprintOpen access

  ABSTRACT Prosocial behaviours—actions that incur personal costs to benefit others—are central to human social life. Two key domains are moral harm aversion, where individuals forgo personal gains to prevent harming others, and prosocial effort, which involves exerting effort to benefit others. Although previous studies suggest a relationship between these behaviours, it remains unclear whether neural responses in one domain can predict prosocial motivation in another. Here, we tested whether neural sensitivity to morally salient information in harm aversion could predict prosocial effort later. Participants completed two tasks: a harm aversion task during fMRI, in which they traded off monetary profit against delivering electric shocks to another person; and, one week later, a prosocial effort task outside the scanner, in which they decided whether rewards for others were worth the required physical effort. We focused on three regions implicated in cost–benefit decision-making and social cognition: the anterior cingulate cortex (ACC), anterior insula (AI), and temporoparietal junction (TPJ). Behaviourally, greater harm aversion was associated with increased prosocial effort. Neurally, AI responses to others’ harm predicted sensitivity to others’ rewards in the effort task, consistent with a role in representing others’ outcomes across positive and negative valences. By contrast, TPJ responses to profit from harming others predicted decreased sensitivity to others’ rewards, suggesting a role in context-dependent valuation that may constrain prosocial behaviour. These findings demonstrate that neural responses to morally salient information in one context correlate with prosocial motivation in another, highlighting mechanisms that bridge moral sensitivity and effortful prosociality.

  PublisherOA PDFDOI

• The pitfalls of pay-to-play morality

  2025-08-29

  articleOpen accessSenior author

  Morality is often measured in dollars and cents. Economic games and charitable giving tasks frame giving money to strangers as moral and keeping money as selfish. This ‘pay-to-play’ paradigm has grown popular with the rise of online experiments, yet its underlying assumptions often go untested. Here, we present data that complicate the interpretation of this paradigm. We invited US crowdworkers (N=1384) to make charitable decisions and measured not just their choices, but the motives and meanings they ascribed to those choices. We found that crowdworkers’ most common motive for keeping money was acute financial need (e.g., struggling to pay for medical care or groceries). These individuals reported significantly lower incomes, and did not view their decisions to keep money as selfish. Third-party judges (N=140) agreed that keeping money in economic tasks out of financial need was not selfish, and considered such tasks to be an unreasonable measure of altruism and selfishness. Finally, we found that crowdworkers reporting financial need were unresponsive to standard interventions designed to motivate generosity. Overall, these findings challenge the assumption that keeping money uniformly reflects selfishness, and raise ethical and practical questions about how experimenters should define, measure, and motivate moral behavior.

  PublisherOA PDFDOI

• Modern maxims for an AI oracle

  Nature Machine Intelligence · 2025-01-13 · 2 citations

  article1st authorCorresponding
  PublisherDOI

• AI Surrogates and Illusions of Generalizability in Cognitive Science Private Project Public Project

  2025-10-23

  articleOpen access1st authorCorresponding

  Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new knowledge about human cognition and behavior. This vision of ‘AI Surrogates’ promises to enhance research in cognitive science by addressing longstanding challenges to the generalizability of human subjects research. AI Surrogates are envisioned as expanding the diversity of populations and contexts that we can feasibly study with the tools of cognitive science. Here, we caution that investing in AI Surrogates risks entrenching research practices that narrow the scope of cognitive science research, perpetuating ‘illusions of generalizability’ where we believe our findings are more generalizable than they actually are. Taking the vision of AI Surrogates seriously helps illuminate a path toward a more inclusive cognitive science.

  PublisherDOI

• How Gendered Moral Norms Amplify Punishment for Selfish Women

  SSRN Electronic Journal · 2025-01-01

  preprintOpen accessSenior author
  PublisherDOI

• Disagreements About Threats to Electoral Integrity: Beliefs About the Severity and Frequency of Fraudulent, Uncounted, and Forgone Votes in the 2020 and 2024 Elections

  Political Behavior · 2025-07-30 · 2 citations

  articleSenior author
  PublisherDOI

• How to show that a cruel prank is worse than a war crime: Shifting scales and missing benchmarks in the study of moral judgment

  Cognition · 2025-09-27 · 3 citations

  article
  PublisherDOI

Recent grants

• Reinforcement Learning Theory in the Digital Age

  NSF · $165k · 2018–2020

Frequent coauthors

• Jenifer Z. Siegel

  Columbia University

  59 shared

• Nadira S. Faber

  38 shared

• Raymond J. Dolan

  National Hospital for Neurology and Neurosurgery

  36 shared

• Patricia L. Lockwood

  University of Birmingham

  34 shared

• Julian Savulescu

  Wellcome Centre for Ethics and Humanities

  34 shared

• William J. Brady

  33 shared

• Jim A. C. Everett

  University of Kent

  32 shared

• Matthew A. J. Apps

  30 shared

Labs

• Crockett LabPI

Awards & honors

• Major prize from National Academy of Science