About

Professor Geoffrey Cohen is a faculty member at the Stanford Graduate School of Education, with additional appointments as a Professor of Psychology and a courtesy Professor of Organizational Behavior. His research examines processes that shape people's sense of belonging and self, focusing on the implications for social problems. He studies the threats to belonging and self-integrity encountered in school, work, and health care settings, and develops strategies to create more inclusive spaces for individuals from diverse backgrounds. Cohen believes that the development of psychological theory is facilitated not only by descriptive and observational research but also by theory-driven intervention, inspired by Kurt Lewin's quote that understanding is best achieved through trying to change it. He is the author of a forthcoming book titled 'Belonging: The Science of Creating Connection and Bridging Divides,' which explores the causes and consequences of a sense of belonging across various social arenas. His work is also supported by affiliations with the Institute for Human-Centered Artificial Intelligence (HAI), the Wu Tsai Human Performance Alliance, and the Stanford Accelerator for Learning.

Research topics

• Political Science
• Psychology
• Computer Science
• Sociology
• Social psychology
• Social Science
• Psychiatry
• Artificial Intelligence
• Business
• Gerontology
• Medicine
• Medical education
• Philosophy
• Knowledge management
• Economics
• Marketing
• Epistemology
• Public relations
• Law
• Pedagogy
• Mathematics education
• Management
• World Wide Web
• Engineering

Selected publications

• Trust in advisor predicts Ph.D. students' academic motivation, well-being, and achievement: A prospective longitudinal study

  PNAS Nexus · 2026-01-01 · 1 citations

  articleOpen access

  Ph.D. training across academic disciplines has faced significant challenges, with news and media outlets regularly highlighting student struggles, including frequent distrust and antagonism in relationships with faculty advisors. Given the centrality of the advisor-advisee relationship in doctoral education-most programs follow an apprenticeship model where students are mentored closely by a primary faculty advisor-the quality of this relationship may be particularly important. Yet almost no empirical research has systematically examined the long-term implications of advisor trust in doctoral education. In a prospective longitudinal study of 558 incoming Ph.D. students-primarily in STEM (Science, Technology, Engineering, and Mathematics) fields-at three US research universities, we examined the role of Ph.D. students' trust in their advisor early in graduate school on 17 outcomes measured repeatedly during students' first year, a sensitive period when a disproportionate number of students have historically exited. Results revealed a surprisingly broad and consistent effect of advisor trust, such that students who had greater trust in their advisor finished their first year more motivated, higher in well-being, and more academically successful than those with lower advisor trust. These effects were independent of student and advisor demographics, students' academic preparation, individual characteristics, and existing differences in outcomes measured at the start of graduate school, highlighting the potential causal implications of advisor trust for shaping a successful and healthy Ph.D. journey.

  PublisherDOI

• Improving social belonging, meaning, and mental health during COVID-19: A self-affirmation approach.

  Emotion · 2025-02-18 · 2 citations

  articleSenior author

  The topics of social belonging, meaning and purpose in life, and mental health have enduring significance, and our objective was to assess the efficacy of two values-affirmation (VA)-based interventions in forestalling critical psychological costs of the COVID-19 pandemic. Both were based on self-affirmation theory (Steele, 1988) and sought to help people stay connected with their core values during this stressful time. One intervention was a one-time VA activity in which participants wrote about important values and were reminded of the value of focusing on core values during uncertain times. A second, elaborated version further guided participants to incorporate brief, values-aligned daily activities. A longitudinal randomized controlled experiment conducted in the United States and Italy revealed an upward trend in social belonging and mental health among participants in the VA conditions-a surprising and positive outcome during a time of looming fear. The trajectories of social belonging and meaning were altered in the intervention conditions, yielding immediate benefits, while long-term benefits (4 weeks postintervention) on social belonging and mental health were confined to men, who experienced poorer psychological outcomes as the pandemic progressed. Additionally, socioeconomic status moderated intervention effects on social belonging, primarily benefitting participants of lower socioeconomic status. Surprisingly, culture and other risk factors (e.g., financial impact of COVID-19; living alone) did not moderate intervention effects. Discussion centers on how tailored VAs can interrupt a recursive cycle triggered by threats to self-integrity, and the potential of social psychologically informed interventions for enhancing belonging, meaning and mental health in face of acute stressors. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

  PublisherDOI

• Understanding the effects of values affirmation on deaf students in Nepal: evidence from a randomized controlled trial

  Social Psychology of Education · 2025-07-04 · 1 citations

  articleOpen accessSenior author

  In this paper, we present two studies from the first randomized controlled trial on the impact of a values affirmation intervention on academic performance and psychological well-being in school for deaf students in secondary deaf schools in Nepal. In Study 1, we found moderate and suggestive evidence that values affirmation had a small effect on lowering students’ perceived stereotype threats and improving their academic performance. Contrary to the evidence in past studies, the affirmation effect was stronger among students who initially reported facing fewer stereotype threats. These findings suggested that the intervention might not counter severe threats for some students but could help those facing objectively high, yet relatively lower, threats by providing a psychological buffer. In Study 2, we conducted a qualitative thematic analysis of students’ reflections in both the treatment and control exercises and used the thematic codes to predict academic performance separately within each treatment condition. We found strong evidence that reflections on the themes “Self-awareness and growth in self-identity” and “Relationship with friends” were positively associated with better academic performance for students in the treatment group. We found no evidence that any themes in the control group negatively predicted academic performance for students in the control group. These findings pinpoint plausible context-speific mechanisms for values affirmation and provide an example research agenda for future values affirmation studies targeting vulnerable populations in novel contexts.

  PublisherOA PDFDOI

• Education

  2025-05-01

  book-chapter1st authorCorresponding

  Education serves as a bridge to the past, preserving a group’s history and identity, while equipping each generation to improve upon the past and adapt to novel circumstances ... Without an effective system of education, each generation would have to reinvent their group’s wisdom, technology, practical know-how, and much more.

  PublisherDOI

• Interprofessional Teamwork Reduces Medical Errors and Burnout: A Multicenter Field Experiment

  Academy of Management Proceedings · 2025-07-01

  articleSenior author

  We argue that both psychological and structural barriers must be overcome to enable interprofessional teamwork, which is crucial for patient safety and clinician well-being. Since nurses are more likely to share information with their peers rather than with doctors, we conducted a values-affirmation intervention to address nurses’ psychological barriers. Additionally, because nurses often lack formal opportunities to fully interact with doctors during patient care, we implemented a multidisciplinary round intervention to address structural barriers. Using a factorial design, we randomly assigned 519 medical teams (1,082 nurses and 847 doctors) to one of four conditions: either intervention alone, both interventions combined, or a control group. Our findings indicate that either intervention significantly reduced doctors’ medical errors via increased information sharing by nurses and decreased nurse burnout through enhanced emotional support from doctors. Larger and more sustainable improvements were observed when the interventions were combined. Manipulation checks showed that nurses were less anxious about participating in multidisciplinary rounds and spoke up more during these rounds after receiving the affirmation intervention. Overall, we demonstrate a synergistic effect of psychological and structural interventions in healthcare, with theoretical implications for integrating behavioral and structural change efforts.

  PublisherDOI

• Ubiquitous Well-being: Patchable AI Unites Computing and Psychological Science

  2025-10-12

  articleSenior author

  Brief psychological interventions (e.g., values affirmation) improve health and learning, yet an infrastructure for personalized, scalable, and timely delivery is missing. We propose that ubiquitous computing can transform static interventions into evolving services. This work introduces Patchable AI, a framework in which chatbots become continuously learning systems through hot-swappable, auditable ''intervention patches'' guarded by safety policies. Four architectural layers—sensor triggers, policy guardrails, patch registry, and conversation broker—allow evolution without retraining underlying models. To provide materials for brief psychological interventions, we conducted a large-scale data collection effort using a custom web-scriber to capture nearly all publicly available psychological and journaling prompts (N = 54,251) from 2015 onward, yielding short, crisp contents structured for patchable deployment. Our values-affirmation case study demonstrates how Patchable AI aligns psychological theory, organizational policy, and technical affordances, offering a roadmap for scalable, personalized well-being services. This research aims to (1) apply a ubiquitous-computing framework across disciplines—education, public health, workplace training—and (2) advance intervention science by embedding human-centered AI that turns static protocols into adaptive, context-aware services. We also outline broader implications, showing how Patchable AI can underpin living support systems across domains.

  PublisherDOI

• Some Surviving, Others Thriving: Inequality in Loss and Coping During the Pandemic

  RSF The Russell Sage Foundation Journal of the Social Sciences · 2024-08-26 · 6 citations

  articleOpen access

  We investigate the contrasting realities of the pandemic on psychosocial experiences and ways of coping among American Voices Project respondent surveys (<i>N</i> = 720) and interviews (<i>N</i> = 172). Despite similar levels of distress early in the pandemic, by late 2020 clear differences across education, race and ethnicity, and gender emerged, both quantitatively and qualitatively. Those with structural advantages reported greater gains from the pandemic, including self-improvement opportunities like therapy and time outdoors. In contrast, respondents without college degrees, Black and Hispanic individuals, and women reported experiencing greater psychosocial shocks into the later months of 2020 and feeling disproportionately undervalued, socially disconnected, and stressed, respectively. The former two groups also systematically differed in their coping strategies, which included hard work, emotion suppression, and faith.

  PublisherOA PDFDOI

• Challenging the N-Heuristic: Effect size, not sample size, predicts the replicability of psychological science

  PLoS ONE · 2024-08-23 · 6 citations

  articleOpen accessSenior authorCorresponding

  Large sample size (N) is seen as a key criterion in judging the replicability of psychological research, a phenomenon we refer to as the N-Heuristic. This heuristic has led to the incentivization of fast, online, non-behavioral studies-to the potential detriment of psychological science. While large N should in principle increase statistical power and thus the replicability of effects, in practice it may not. Large-N studies may have other attributes that undercut their power or validity. Consolidating data from all systematic, large-scale attempts at replication (N = 307 original-replication study pairs), we find that the original study's sample size did not predict its likelihood of being replicated (rs = -0.02, p = 0.741), even with study design and research area controlled. By contrast, effect size emerged as a substantial predictor (rs = 0.21, p < 0.001), which held regardless of the study's sample size. N may be a poor predictor of replicability because studies with larger N investigated smaller effects (rs = -0.49, p < 0.001). Contrary to these results, a survey of 215 professional psychologists, presenting them with a comprehensive list of methodological criteria, found sample size to be rated as the most important criterion in judging a study's replicability. Our findings strike a cautionary note with respect to the prioritization of large N in judging the replicability of psychological science.

  PublisherDOI

• When politics trumps truth: Political concordance versus veracity as a determinant of believing, sharing, and recalling the news.

  Journal of Experimental Psychology General · 2024-10-01 · 6 citations

  articleSenior author

  Resistance to truth and susceptibility to falsehood threaten democracies around the globe. The present research assesses the magnitude, manifestations, and predictors of these phenomena, while addressing methodological concerns in past research. We conducted a preregistered study with a split-sample design (discovery sample N = 630, validation sample N = 1,100) of U.S. Census-matched online adults. Proponents and opponents of 2020 U.S. presidential candidate Donald Trump were presented with fake and real political headlines ahead of the election. The political concordance of the headlines determined participants' belief in and intention to share news more than the truth of the headlines. This "concordance-over-truth" bias persisted across education levels, analytic reasoning ability, and partisan groups, with some evidence of a stronger effect among Trump supporters. Resistance to true news was stronger than susceptibility to fake news. The most robust predictors of the bias were participants' belief in the relative objectivity of their political side, extreme views about Trump, and the extent of their one-sided media consumption. Interestingly, participants stronger in analytic reasoning, measured with the Cognitive Reflection Task, were more accurate in discerning real from fake headlines when accurate conclusions aligned with their ideology. Finally, participants remembered fake headlines more than real ones regardless of the political concordance of the news story. Discussion explores why the concordance-over-truth bias observed in our study is more pronounced than previous research suggests, and examines its causes, consequences, and potential remedies. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

  PublisherDOI

• Investigating Cortisol in a STEM Classroom: The Association Between Cortisol and Academic Performance

  Personality and Social Psychology Bulletin · 2023-08-02 · 3 citations

  article

  Science, technology, engineering, and mathematics (STEM) education can be stressful, but uncertainty exists about (a) whether stressful academic settings elevate cortisol, particularly among students from underrepresented racial/ethnic groups, and (b) whether cortisol responses are associated with academic performance. In four classes around the first exam in a gateway college STEM course, we investigated participants’ ( N = 271) cortisol levels as a function of race/ethnicity and tested whether cortisol responses predicted students’ performance. Regardless of race/ethnicity, students’ cortisol, on average, declined from the beginning to the end of each class and across the four classes. Among underrepresented minority (URM) students, higher cortisol responses predicted better performance and a lower likelihood of dropping the course. Among non-URM students, there were no such associations. For URM students, lower cortisol responses may have indicated disengagement, whereas higher cortisol responses may have indicated striving. The implication of cortisol responses can depend on how members of a group experience an environment.

  PublisherDOI

Recent grants

• Reducing the Racial Achievement Gap: A Self-Affirmation Intervention Approach

  NSF · $948k · 2007–2011

• Understanding Peer Influence of Adolescent Health Risk Behaviors

  NIH · $2.5M · 2009–2015

Frequent coauthors

• Julio Garcia

  Stanford University

  34 shared

• Valerie Purdie‐Vaughns

  Columbia University

  27 shared

• John F. Steiner

  British Psychoanalytical Society

  25 shared

• L. Miriam Dickinson

  University of Colorado Denver

  25 shared

• Irene V. Blair

  25 shared

• Channing E. Tate

  University of Colorado Denver

  25 shared

• Rebecca Hanratty

  Denver Health Medical Center

  25 shared

• Nancy H. Apfel

  Yale University

  22 shared

Labs

• Cohen LabPI