About

Daniel Epstein is a faculty member at the Connected Learning Lab, where he applies Human-Computer Interaction methods to study personal tracking technology. His research focuses on how these tools can better acknowledge and account for the realities of everyday life. Through his work, he contributes recommendations on designing tracking tools that are more inclusive of people's identities, environments, and abilities.

Research topics

• Computer Science
• Political Science
• Applied psychology
• Psychology
• Social Science
• Internet privacy
• Sociology
• Medicine
• Social psychology
• Psychiatry
• Knowledge management
• Data science
• Public relations
• Clinical psychology
• World Wide Web
• Nursing

Selected publications

• CASEbot: A Conversational Agent for Structuring and Personalizing the Design of Self-Experiments in Personal Health

  2026-04-13 · 1 citations

  articleOpen access

  Self-experimentation, or using tracked data to systematically answer health and wellbeing questions via hypothesis testing, has significant potential to support personal health. However, technological support for self-experimentation has focused on expert-designed self-experiments for specific health conditions, limiting people’s ability to design their own rigorous experiments. To address this gap, we developed CASEbot (Conversation Agent for Self-Experimentation), an LLM-powered chatbot using a theory-driven approach to guide users through designing well-structured, personalized, and safe self-experiments. We conducted a within-subjects, mixed-methods study with 42 participants comparing CASEbot to a traditional worksheet-based approach. When formally comparing the experiment rigor and specificity, most participants designed better experiments using CASEbot. They appreciated CASEbot’s conversational approach, which prompted them to surface everyday constraints and proactively raised safety concerns, but some found the platform too rigid in its recommendations. We discuss opportunities for future generative AI self-experimentation systems for health to balance structured guidance with user autonomy.

  PublisherDOI

• Conferences: What's in Scope for Interactive Health?

  Voices of SIGCHI · 2026-04-01

  article1st authorCorresponding

  Over the last decade, Health has emerged as a vibrant area of focus within HCI. Since 2024, the Health Subcommittee has grown by nearly 30% annually, making it one of the largest subcommittees with over 500 papers being processed for CHI 2026. This growth has led to multiple attempts to found a Health and HCI conference, and finally in July 2026 we will celebrate the efforts of the community in the inaugural SIGCHI sponsored Interactive Health Conference in Porto, Portugal.

  PublisherDOI

• Understanding Adoption, Use, and Abandonment Practices in Baby Tracking

  2026-04-13 · 1 citations

  articleOpen accessSenior author

  New parents often turn to baby-tracking technology to monitor and reflect on the daily routines of their infants. However, we lack understanding of how tracking practices evolve as children grow and develop, with caregivers adopting, using, and eventually abandoning baby tracking. We analyze the logs of 60 parents and 71 children who used the popular baby-tracking app Huckleberry for an average of 12 months, combined with re-analyzing interviews with 20 parents who used various baby-tracking technologies. We find that parents start tracking at different ages, track habitually and intermittently, change and swap what and how they track, and often gradually abandon the practice. Through unpacking why these patterns occur, we find that parents effectively self-manage what data categories are worthwhile to continue tracking. We point out lessons that domains outside baby tracking can take from the evolving, longitudinal process, and present design recommendations to better support caregivers across phases.

  PublisherDOI

• Unpacking How Pole Dancers Come to Experience the Use of Fitness Trackers

  2026-04-13 · 1 citations

  articleOpen accessSenior author

  Personal informatics literature has typically examined how fitness trackers can support understanding of cardio-based exercises, like running, which are closely associated with the step counts central to these devices. Although this focus has supported individuals in gleaning fruitful data discoveries, it confines the manner in which a body moves to these movement domains. In this paper, we contend with pole dance, a physical activity whose movement affordances (e.g., rotational, artistic, feminine) differ greatly from those centralized in the self-tracking technologies. We qualitatively interviewed 20 polers, and gathered their reflection on pole dance and how they view the use self-tracking devices for the activity. Using their insights on the inherent dynamicity of poling, independently, and upon interaction with fitness trackers, we offer suggestions to our discipline to challenge ourselves to re-imagine sensing as a pluralistic endeavor.

  PublisherDOI

• FamilyBloom: Examining Ecologies of Collaboration in Family-Centered Health Tracking

  2026-04-13 · 1 citations

  articleOpen accessSenior author

  Family health informatics tools can help support well-being with shared data tracking. Prior work typically focused on shared data review, but often in specific moments, like bedtime, or centered on caregiving of children or elderly members. To investigate how tracking can support mutual health collaboration between family members pervasively across daily contexts, we designed and deployed FamilyBloom, a glanceable smartwatch and home display system for mood and goal tracking. Twelve families with both neurotypical and ADHD members used FamilyBloom for three months on average. Our findings reveal how family-centered tracking created collaboration opportunities and tensions across multiple ecological systems: individual self-regulation, collaborations within family dynamics, involvement of care networks with varying trust levels, institutional school constraints and cultural stigma, and temporality of regular routines and crisis periods. We discuss an ecosystem-aware approach to family informatics, wherein design can attend to how families navigate multiple contexts while sustaining family-level collaboration.

  PublisherOA PDFDOI

• Sharing Women’s Health Experiences Influenced by Genetic Factors: Practices, Challenges, and Opportunities

  2026-04-13 · 1 citations

  article
  PublisherDOI

• Mycobacterium Parascrofulaceum Infection Mimicking Bronchiolitis Obliterans Syndrome: The Value of Quantitative Chest CT Analyzed With Parametric Response Mapping

  American Journal of Respiratory and Critical Care Medicine · 2025-05-01

  article

  Abstract Introduction: Mycobacterium parascrofulaceum is an extremely rare opportunistic pathogen. The presentation of M. parascrofulaceum infections can overlap with other pulmonary diseases, including bronchiolitis obliterans syndrome (BOS), a rare and severe complication of allogeneic hematopoietic cell transplant (HCT). Infections and graft-versus-host-disease (GVHD) can present similarly, leading to delays in diagnosis and poor outcomes. Quantitative computed tomography (qCT) scans, involving paired images in the respiratory cycle, provide radiographic and physiologic information and offer unique diagnostic insights for pulmonary complications after HCT. Case Description: A 54-year-old female with a history of breast cancer and therapy-related acute myeloid leukemia underwent an allogeneic HCT. Eight months later, she presented with dyspnea and cough. She was initially diagnosed with infection due to parainfluenza type 3. A chest CT showed multi-lobar ground-glass opacities suggestive of atypical or viral infection or organizing pneumonia (OP). She was treated with methylprednisolone (1 mg/kg/day) for presumed viral-induced OP, with rapid improvement in her dyspnea. A subsequent CT scan demonstrated radiographic improvement, and her corticosteroids were tapered. Shortly after corticosteroids were discontinued, she experienced worsening dyspnea. A CT angiogram of her chest revealed a moderate pericardial effusion. Prednisone (1 mg/kg/day) was started for chronic GVHD (manifesting as serositis), and subsequent echocardiography demonstrated resolution of her pericardial effusion. She remained asymptomatic but her forced expiratory volume (FEV1) in 1 second, percent predicted continued to decline, from 95% pre-transplant to 58%, raising concern for BOS. A qCT scan of the chest with parametric response mapping (PRM) analysis demonstrated tree-in-bud nodules concerning for infection but was not consistent with BOS given no significant air trapping (Figure 1). Two of three induced sputum specimens sent for acid-fast culture were smear-negative but both grew M. parascrofulaceum. Novelty and Importance: This case demonstrates how a nontuberculous mycobacterial infection can result in FEV1 decline mimicking BOS. Moreover, it demonstrates the diagnostic value of qCT with PRM analysis, which can both rule out significant air trapping and identify radiographic features consistent with infection like tree-in-bud nodules. Results of this tool were used to correctly identify and treat an infection and avoid consequences of inappropriate immunosuppression.

  PublisherDOI

• Family Dynamics on Data Tracking and Sharing with Children-Oriented Wearables

  SSRN Electronic Journal · 2025-01-01

  preprintOpen accessSenior author
  PublisherDOI

• Foody Talk: Exploring Opportunities for Conversational Food Journaling

  2025-04-24 · 4 citations

  articleOpen accessSenior author
  PublisherDOI

• Understanding How Personal Activities Are Shared In Short-form Videos

  Proceedings of the ACM on Human-Computer Interaction · 2025-10-16 · 1 citations

  articleOpen accessSenior author

  Sharing activities that people do in everyday life, such as physical activity, health management, or hobbies, help people receive benefits like social support and positive self-presentation. Short-form videos present new opportunities for activity-sharing, which has traditionally been studied in static contexts like text- and image-sharing. We therefore aim to understand what information people incorporate into short-form activity videos, and how. We qualitatively analyzed 420 short-form activity videos on TikTok across three domains: running, studying, and sketching. We found people often present information before, during, and after activities, developing strategies for qualitatively and quantitatively incorporating activity-relevant information in each. We also uncover practices for aligning the sharing of activity-relevant information with the nature of short-form videos, such as modifying broader-scale goals into video-scale goals. We further discuss design opportunities and challenges for designers to create tools that support the practice, such as closer integration with tracking tools and encouraging narrative structure.

  PublisherDOI

Recent grants

• CRII: CHS: Examining a Multimodal Approach to Lowering the Burden of Food Journaling

  NSF · $175k · 2019–2022

Frequent coauthors

• Eliseo Berní Reategui

  20 shared

• Sean A. Munson

  14 shared

• James Fogarty

  University of Washington

  13 shared

• Eunkyung Jo

  University of California System

  10 shared

• Isis da Costa Pinho

  Universidade Federal do Rio Grande do Sul

  10 shared

• Nasser Assadi

  Regional Medical Center

  9 shared

• Shawn Kurian

  9 shared

• Chulou H Penales

  Nova Southeastern University

  9 shared

Education

• Ph.D., Computer Science & Engineering

  University of Washington

  2018

• B.S., Computer Science

  University of Virginia

  2012