About

Sean Young is a social/behavioral psychologist and health services researcher. He studies how to use social technologies to change (social, digital interventions) and predict (data science/AI) health and societal problems. He is a social entrepreneur and the author of the #1 Wall Street Journal/international best-selling book, Stick with It, on the science behind lasting behavior change. Sean Young is an Associate Professor of Emergency Medicine and Informatics at the Schools of Medicine and ICS.

Research topics

• Medicine
• Political Science
• Nursing
• Psychiatry
• Psychology
• Applied psychology

Selected publications

• Drug and Substance Abuse Point of Interest (POI) conflation and spatial attribute enrichment framework

  Figshare · 2026-03-17

  articleOpen accessSenior author

  Point of Interest (POI) data is essential for accurate spatial analysis of pressing healthcare issues such as Drug and Substance Abuse (DSA). However, retrieving accurate healthcare POI information remains complicated. POI conflation integrates data from multiple sources to enhance spatial attribute quality and coverage. This study proposes a multi-step framework for healthcare POI conflation, including POI data collection from Location-Based Services (LBS), geographic and spatial attributes collection, calculating similarity across datasets, POI matching, spatial attributes enrichment, manual labeling, quality control, and deployment. We tested this framework on a DSA use case in California, USA. Our automated approach was able to detect 11,936 unique POIs related to this healthcare tag. Of the locations used for use case validation, 33% were common with the commercial benchmark POI dataset used by Safegraph. Out of the 11,936 total POIs, we were only able to find 4535 (38%) that were not in the commercial benchmark dataset and relevant to the use case. Moreover, our results indicate that spatial attributes fill rates are 32% at the geometry building level and 98% at the Census block group (CBG) level. We conclude that using LBS can provide POIs with relevance and spatial attributes similar to commercial datasets. Public health agencies should integrate LBS data to enhance the accessibility, accuracy, and timeliness of healthcare POI information.A multi-source POI conflation framework improves spatial analytics by combining public, commercial, and volunteered datasets.Accurate spatial attributes, such as Census Block Group and building geometry, enable more targeted demographic and mobility-based health interventions. Public health agencies should integrate LBS data to enhance the accessibility, accuracy, and timeliness of healthcare POI information. A multi-source POI conflation framework improves spatial analytics by combining public, commercial, and volunteered datasets. Accurate spatial attributes, such as Census Block Group and building geometry, enable more targeted demographic and mobility-based health interventions.

  PublisherDOI

• Drug and Substance Abuse Point of Interest (POI) conflation and spatial attribute enrichment framework

  Cartography and Geographic Information Science · 2026-03-17

  articleSenior authorCorresponding
  PublisherDOI

• Disparities in HIV Testing Among Spanish and English-Speaking Hispanic/Latino Men who have Sex with Men and Social Media Use

  AIDS and Behavior · 2026-05-20

  articleSenior author
  PublisherDOI

• Drug and Substance Abuse Point of Interest (POI) conflation and spatial attribute enrichment framework

  Figshare · 2026-03-17

  articleOpen accessSenior author

  Point of Interest (POI) data is essential for accurate spatial analysis of pressing healthcare issues such as Drug and Substance Abuse (DSA). However, retrieving accurate healthcare POI information remains complicated. POI conflation integrates data from multiple sources to enhance spatial attribute quality and coverage. This study proposes a multi-step framework for healthcare POI conflation, including POI data collection from Location-Based Services (LBS), geographic and spatial attributes collection, calculating similarity across datasets, POI matching, spatial attributes enrichment, manual labeling, quality control, and deployment. We tested this framework on a DSA use case in California, USA. Our automated approach was able to detect 11,936 unique POIs related to this healthcare tag. Of the locations used for use case validation, 33% were common with the commercial benchmark POI dataset used by Safegraph. Out of the 11,936 total POIs, we were only able to find 4535 (38%) that were not in the commercial benchmark dataset and relevant to the use case. Moreover, our results indicate that spatial attributes fill rates are 32% at the geometry building level and 98% at the Census block group (CBG) level. We conclude that using LBS can provide POIs with relevance and spatial attributes similar to commercial datasets. Public health agencies should integrate LBS data to enhance the accessibility, accuracy, and timeliness of healthcare POI information.A multi-source POI conflation framework improves spatial analytics by combining public, commercial, and volunteered datasets.Accurate spatial attributes, such as Census Block Group and building geometry, enable more targeted demographic and mobility-based health interventions. Public health agencies should integrate LBS data to enhance the accessibility, accuracy, and timeliness of healthcare POI information. A multi-source POI conflation framework improves spatial analytics by combining public, commercial, and volunteered datasets. Accurate spatial attributes, such as Census Block Group and building geometry, enable more targeted demographic and mobility-based health interventions.

  PublisherDOI

• Can People Tell the Difference Between AI-Generated Mental Health Vignettes? An Exploratory Comparison of User Evaluations

  Information · 2026-02-12

  articleOpen accessSenior author

  Vignettes are brief, descriptive, hypothetical scenarios that have been used to extract attitudes, beliefs, or perceptions from participants across psychology, healthcare, and human–computer interaction. Traditional vignette development is often time and labor-intensive and large language models (LLMs) like ChatGPT-4o may streamline this process. This exploratory between-subjects online survey (n = 66) compared participants’ perceptions of clinically reviewed LLM-generated versus human-written mental health vignettes describing social anxiety, depression, or schizophrenia. Participants rated each vignette on realism, clarity, engagement, emotional impact, perceived likelihood of AI authorship, and likelihood that the target diagnosis applied. Mixed-effects linear regression analyses showed no statistically significant differences between AI-generated and human-written vignettes for any perceived quality rating; estimated source effects were small (|β| ≤ 0.10) with 95% confidence intervals spanning zero across outcomes. Perceived AI authorship likelihood (β = 0.09, 95% CI [−0.22, 0.40]) and correct-diagnosis likelihood ratings (β = −0.07, 95% CI [−0.30, 0.16]) also did not differ by source. Overall, we did not detect statistically significant differences between AI-generated and human-written vignettes. These findings reflect perceptions of AI-generated vignettes that underwent expert clinical review and suggest that LLMs may assist in vignette generation with expert oversight, while highlighting the need for further research on clinical accuracy, diagnostic validity, and generalizability.

  PublisherDOI

• Social media interventions and the moderation of baseline substance use: A secondary data analysis

  Addictive Behaviors · 2026-01-08

  articleOpen access
  PublisherOA PDFDOI

• Enhancing PFAS data integrity: An LLM-based FAIR+Environmental principle for improved evaluation of environmental contaminants and related constituent databases

  Journal of Hazardous Materials Advances · 2026-02-28

  articleOpen access

  • PFAS datasets assessed with FAIR principles and data quality, integration metrics. • A semi-automated LLM assessment pipeline reliably assessed >100 PFAS datasets. • PFOA in groundwater is higher compared to surface, drinking water. • ∼34% of drinking water samples exceed PFOA MCL (4 ng/L). • Public PFAS soil occurrence datasets are limited. Per- and polyfluoroalkyl substances (PFAS) are persistent, bioaccumulative contaminants of emerging concern, yet data sharing around their environmental occurrence and monitoring remains fragmented. We propose a “FAIR+Environmental” framework that extends the Findable, Accessible, Interoperable, Reusable (FAIR) principles to environmental-specific matrices, assessed by a semi-automated Large Language Model (LLM) pipeline that uses rule-based scoring, few-shot prompting, and Chain-of-Thought (CoT) reasoning. We applied the framework to >100 U.S. PFAS datasets across groundwater, surface water, drinking water, and soil matrices. Few-shot CoT LLMs streamlined FAIR evaluations, reducing the manual effort required for expert assessments. Among environmental matrices, surface water datasets achieved the highest FAIR-Score (53.6%), followed by drinking water (52.6%), soil (49.3%), and groundwater (45.2%). Multi-state datasets consistently outperformed single-state datasets, particularly in Interoperability and Reusability criteria. Compared to geoscience databases, PFAS environmental datasets lag in FAIR adherence, highlighting the urgent need for centralized, standardized, and FAIR-compliant data management in the field. PFOA occurrence indicated overall PFAS pollution hotspots because of its frequent detection. PFOA contamination was most severe in soil and groundwater. Surface water and drinking water showed lower concentrations but remain critical public health exposures, with ∼34% of drinking water samples exceeding the 4 ng/L maximum contaminant level. Temporal trends indicated little significant change in PFOA concentrations across most states.

  PublisherDOI

• Artificial intelligence in HIV research, policy, and clinical care

  The Lancet HIV · 2025-10-29

  review1st authorCorresponding
  PublisherDOI

• Mindful Jazz and Preferred Music Interventions Reduce Pain Among Patients With Chronic Pain and Anxiety: A Pilot Randomized Controlled Trial

  Cureus · 2025-03-12 · 1 citations

  articleOpen access1st authorCorresponding

  Background: A mindfulness-based intervention (MBI) focused on listening to music might reduce chronic pain and provide a new approach to overcoming challenges from traditional MBIs (e.g., breathing). Due to the potential unpredictability and unfamiliarity of jazz, an MBI focused on listening to improvisational jazz music might be a particularly efficacious pain reduction intervention. This pilot study explores whether mindfully listening to music, including jazz, can reduce pain-related outcomes. Methods: Chronic musculoskeletal pain (CMP) participants (n=30 per group, N = 120 total) were enrolled online between 12/7/2023 and 2/8/2024. Participants were randomly assigned to one of four groups in a 2 (Mindful Music Listening/Intervention vs. Music Education/Control) X 2 (Preferred Music (choose their own music genre) vs Jazz (assigned to listen to improvisational jazz)) experiment, for a total of four groups (Mindful Jazz, Mindful Music, Jazz Education, and Music Education). Patients in each group were provided with training in either mindful listening to music (Intervention groups) or music education (Control groups) and given four sets of weekly recordings related to their group for daily listening/practice. Patients completed online surveys on pain-related outcomes (e.g., pain catastrophizing, pain intensity, and anxiety) pre- and post-training (immediate outcomes), and throughout a four-week period (longer-term outcomes). The main outcomes analyses compared the intervention and control groups, with secondary sub-analyses among participants who listened to at least 2/3 of their recordings (10 minutes), and among those who experienced a clinically meaningful (20%) reduction in pain. Results: Mindful Jazz and Mindful Music (Intervention) participants reported significantly less pain intensity (p < 0.001) and pain unpleasantness (p < 0.001) immediately after the training relative to the Jazz Education and Music Education (Control) participants. Mindful Jazz participants also reported a significant reduction in anxiety compared to the Jazz and Music Education groups (p < 0.05). Throughout the four-week period, Mindful Jazz participants reported less pain intensity relative to both control groups (Jazz and Music Education); Mindful Music participants reported significantly less pain intensity relative to only the Jazz Education participants. Mindful Jazz participants reported a >20% decrease in pain intensity more frequently than Jazz Education (Χ2=48.71, p<0.001), Music Education (Χ2=65.13, p<0.001), and Mindful Music (Χ2=8.74, p=0.003) participants. Similarly, among the instances when a participant listened to at least 10 minutes of their audio recording, the proportion who achieved a >20% decrease in pain intensity differed significantly ((Χ2=84.03, p<0.001): Jazz Education, 29%; Music Education, 26%, Mindful Jazz, 50%; Mindful Music 41%). Conclusion: Mindfully listening to music can help to reduce pain-related outcomes. Both music education (i.e., music listening without mindfulness training) and mindfully listening to music (i.e., listening with mindfulness training) helped to decrease pain and anxiety from baseline to follow-up. However, mindful listening reduced pain to a greater amount compared to music education, suggesting that mindfully listening to music is a more impactful pain reduction intervention compared to listening without mindfulness training. Future research is warranted with a larger sample.

  PublisherOA PDFDOI

• Systematic Review of Electronic Monitoring to Increase Medication Adherence in Children With Asthma

  Pediatric Pulmonology · 2025-10-01

  reviewSenior author

  INTRODUCTION: While asthma can be well-controlled with the use of inhaled corticosteroids (ICS), most children do not adhere to treatment, leading to exacerbations, emergency department (ED) visits, hospitalizations, and deaths. Electronic monitoring devices (EMDs) have been shown to increase ICS adherence and improve asthma outcomes, but have not been well-studied in young people. The aim of this systematic review is to determine the feasibility and efficacy of EMDs in children with asthma. METHODS: This review was conducted according to PRISMA guidelines. PubMed and CINAHL were searched using keywords such as asthma, wearable electronic devices, and medication adherence. Search results were screened and appraised using Covidence software. Articles were included if they were in English, peer-reviewed, published since 2000, used EMDs for inhaled corticosteroid (ICS) treatment adherence, and included participants with asthma under age 21. RESULTS: Fourteen studies met inclusion criteria, with durations ranging from 2 to 12 months. Recruitment rates varied from 28% to 100%, and retention rates ranged from 51% to 100%. Participants found EMDs acceptable in four of six and feasible in five of six studies, highlighting ease of use and perceived benefits. Six of the 12 studies measuring ICS adherence reported improvements using EMDs with feedback compared to monitoring alone. Seven out of nine studies found improvements in asthma control, and four out of seven found reduced ED visits and hospitalizations. These findings highlight the overall efficacy of EMDs. CONCLUSION: EMDs can improve ICS adherence in children with asthma, especially with sustained feedback. However, inconsistent results highlight the need for comprehensive approaches addressing behavioral and systemic factors.

  PublisherDOI

Recent grants

• HOPE Social Media Intervention for HIV Testing and Studying Social Networks

  NIH · $2.9M · 2015–2022

• Mining real-time social media big data to monitor HIV: Development and Ethical Issues

  NIH · $2.7M · 2017–2021

• NIH Grant R56AI125105

  NIH · $671k · 2019

• NIH Grant K01MH090884

  NIH · $842k · 2015

• NIH Grant R21DA039458

  NIH · $314k · 2018

Frequent coauthors

• Jeffrey D. Klausner

  University of Southern California

  55 shared

• Robert Marlin

  University of California, San Diego

  41 shared

• Emily Huang

  Methodist Hospital

  41 shared

• Joseph Daniels

  Arizona State University

  41 shared

• A Medline

  38 shared

• Greg Wilson

  38 shared

• Lina Rosengren

  Skåne University Hospital

  37 shared

• William G. Cumberland

  University of California, Los Angeles

  33 shared