Shared motivations, goals and values in the practice of personal science: a community perspective on self-tracking for empirical knowledge

The paper examines personal science—self-research and self-experimentation arising from self-tracking—as an emergent N-of-1 knowledge practice. Building on prior work in personal informatics, patient-led research, and science and technology studies, the authors aim to understand how individuals move beyond passive tracking to generate empirical self-knowledge. The study explores: (1) What intrinsic and extrinsic motivations drive individuals to engage in self-research, and how these evolve over time; and (2) How these individual motivations relate to shared goals, values, and practices within personal science communities. The context includes links to citizen science, scientific ethos (Mertonian norms), and peer production, positioning personal science as potentially a participatory, inclusive form of science.

- Personal informatics and HCI: Research has focused on self-tracking for health and wellness, user needs, barriers, and design implications (e.g., lifelogging, lived informatics). Gaps include supporting interpretation/sensemaking and collaborative aspects of self-tracking. Some studies note issues of rigor in self-research and evolving quantified-self consciousness with self-reflection and data sensitivity. - Patient-led communities: Self-tracking supports understanding and decision-making in chronic conditions (e.g., Parkinson’s, diabetes). Communities co-create knowledge via collective self-experimentation, openness, and peer support, though issues of datafication, power asymmetries, and skepticism from professionals persist. Goal elicitation is crucial to avoid ineffective routines; communities often adopt ethical ideals and rational skepticism. - Social studies of science: Beyond critiques of surveillance, digital divides, and biopolitics, scholars describe how self-trackers attribute meaning through reflective, open-ended relationships with data, becoming expert users. Community contexts (e.g., Quantified Self, Open Humans) facilitate experimentation and learning. Self-tracking aligns with democratization of science, DIY science, and citizen science, though links to individual self-research are underexplored. Personal science involves cycles of questioning, designing, observing, reasoning, and discovering, and shares values with Mertonian norms and peer production.

Study context: The first Keating Memorial (KM) Self-research initiative (Feb–Jul 2020), co-organized by Quantified Self and Open Humans, provided a community setting with weekly self-research chats and an end-of-cycle seminar. Projects included sleep, glucose monitoring, tremors, diet-mood, and noise sensitivity. Design: Qualitative, interpretivist study using semi-structured interviews to explore motivations, goals, and values of personal science practitioners. Participants and recruitment: 22 interviewees. Initial participants were selected based on: (1) having conducted or attempted personal science; (2) attending at least two Open Humans community calls; and/or (3) participating in QS forums or OH Slack. Snowball sampling added participants, including some outside KM and QS/OH communities. Ethical approval obtained; all participants consented. Data collection: Interviews via Google Meet (~1 hour), recorded and auto-transcribed (Tactiq), then manually corrected. GDPR-compliant tools used. Analysis framework: A conceptual framework integrating citizen science motivation studies, personal science inquiry cycles, Mertonian scientific norms, and peer production informed a five-category codebook: (1) Improving personal conditions; (2) Enjoying data/tech/research activity; (3) Extrinsic motivations; (4) Contributing to empirical self-knowledge; (5) Sharing goals and values with peers. Subcategories (16 total) refined iteratively. Coding and reliability: Transcripts coded in Taguette by two researchers (ESH, MO). Subcoding independently assigned; Cohen’s kappa for categories 1–4 was 0.82 (strong agreement). Category 5 yielded kappa 0.22; all authors reached consensus by discussion. BGT adjudicated ties for categories 1–4. Open dataset of coded excerpts available (Zenodo: 10.5281/zenodo.5543445). Sample characteristics: 18 participants from KM/QS; 4 via snowball (3 outside similar communities, 1 only QS). Geography: 14 USA; 3 UK; 1 each from Switzerland, Sweden, Netherlands, Canada, Spain. Majority male (16). Ages ranged from university students to retirees. Several had prior scientific training, often applied outside their trained field.

- Scope and coding: - 269 excerpts coded across five categories. Each interview had at least 3 categories; 18 interviews had 4 of 5 categories coded. Intercoder reliability for categories 1–4: κ=0.82; category 5 κ=0.22 (resolved by consensus). - Initial intrinsic motivations: improving personal conditions - Many began with specific health/well-being goals (e.g., sleep disorders, chronic conditions, diabetes, Parkinson’s). Patient-led motives often expanded to related topics (e.g., adding COVID-19 tracking alongside Parkinson’s). - Participants sought knowledge beyond standard wearable/app metrics; some found app-based tracking tedious and preferred customized or manual methods. - Enjoyment of data, technology, and research activity reinforces engagement - Widespread enjoyment using tools, sensors, and data workflows; an early-adopter or “hacker/DIY” attitude common (e.g., building or appropriating CGM, EEG, blood tests). - Inductive exploration prevalent: participants discovered patterns and correlations before formulating specific hypotheses, which then prompted more structured inquiries. - Open reuse/sharing of tools and approaches within the community motivated continued participation. - Curiosity about oneself and a growth mindset underpinned sustained engagement; both high-tech (wearables) and low-tech (journaling, spreadsheets) methods used. - Extrinsic motivations play a minor role - Some overlap with professional development (skill acquisition, applying tools across work and self-research), occasional interest in academic publication, and rare mentions of business opportunities. Overall, extrinsic drivers were secondary. - Learning and sharing empirical knowledge are common goals - Community processes (talks, discussions, feedback) strongly motivated participants to share work-in-progress, including failures. - Interest in scaling from N-of-1 to “N-of-many” through data aggregation and sharing, especially evident in patient-led contexts (e.g., CGM communities). Sharing often remained unstructured, with variability in protocols beyond data collection. - Shared values aligned with scientific and social practices - Sociality/communality: desire to meet like-minded peers; sharing tools, protocols, and results as common property. - Organized skepticism: peers openly critique protocols, tool reliability, and results in supportive, non-judgmental settings. - Universalism less evident beyond general openness; disinterestedness partially present via willingness to share outputs for communal benefit despite personal starting points. - Additional descriptive findings - Majority of interviewees from the US; diverse ages; many had scientific backgrounds, often applied outside their formal fields.

Findings address the research questions by showing that motivations in personal science are multi-faceted, interrelated, and evolve over time. Initial health and well-being goals catalyze deeper, more rigorous inquiry that extends beyond commercial metrics. Enjoyment of data work and research processes reinforces engagement, and the appropriation of tools (including medical devices) enables exploratory and iterative investigations. Extrinsic motivations—career development, publications, business—are present but generally secondary. Community practices of sharing and learning are central: participants value feedback, transparency, and collectively iterating on methods, fostering organized skepticism and communality analogous to scientific norms and peer production cultures. The study situates personal science within the broader citizen science paradigm: a participatory, inclusive scientific culture driven by critical thinking, openness, and collaboration. Aspirations to scale from N-of-1 to larger aggregates highlight pathways for participant-led knowledge production, alongside challenges in time demands, protocol standardization, and engagement with medical professionals.

The study offers a conceptual and practitioner-informed account of motivations, goals, and values in personal science. It shows that long-term engagement arises from intertwined intrinsic motivations—health/well-being improvement, curiosity, enjoyment of data/research—and community-driven goals of learning and sharing, while extrinsic motivations play a lesser role. The work argues that personal science is a specific form of extreme citizen science, contributing to a more participatory and inclusive scientific culture. Implications include designing personal informatics and patient-led research tools that support sensemaking, curiosity, and community sharing, to improve sustained engagement and benefits. Future research should extend the framework quantitatively and qualitatively across broader populations, refine learning/sharing subcategories, examine pathways for scaling to “N-of-many,” and explore collaborations between personal scientists, clinicians, and academics.

- Generalizability: Qualitative findings from 22 interviews are not fully generalizable beyond the sample. - Framework constraints: The personal science research-cycle stages often did not appear formalized or linear, indicating limitations of the integrative framework; learning/sharing subcategories need further refinement. - Coding reliability: Low intercoder agreement for values-related category (category 5; κ=0.22) required consensus coding. - Practice heterogeneity and unstructured sharing: Variability in protocols beyond data collection and time demands may affect comparability and scalability. - Engagement with professionals: Reported hesitance from clinicians to engage with self-tracking insights indicates contextual barriers.