Epistemic Agency: AI Qualitative Analysis of Self-Tracking

Researchers and practitioners are asking whether self-tracking is merely data production or a form of knowledge creation: the Frontiersin.org review frames this question as epistemic agency in self-tracking and maps a spectrum from low to high agency. The primary keyword for this article is "epistemic agency in self-tracking" and this post shows how AI-enabled qualitative research methods can extract who asked the question, who analyzed the data, and who benefited from the answer. The audience is qualitative researchers, UX and health researchers, and teams running citizen science or patient-led projects who want reproducible syntheses of interviews, self-experiments, and open-ended reports.

Key Takeaways

The Frontiersin.org conceptual review (published 18 August 2026) shows that epistemic agency in self-tracking ranges from participant-as-sensor to participant-as-researcher, and AI-enabled qualitative methods make those differences measurable and actionable. Frontiersin.org

A literature search in April–May 2025 returned 721 records and, after screening, 241 full papers were retained for analysis, according to the Frontiersin.org review (published 18 August 2026).
Of the 241 papers the review analyzed, 46 used self-tracking as the data-collection method and were classified by agency: 17 low agency, 10 mixed agency, and 19 high agency, per Frontiersin.org (18 August 2026).
The review highlights that the key distinction is 'who owns the question': when the individual defines the question, self-tracking becomes personal science (Frontiersin.org, 18 August 2026).

What Happened and How the Review Measured Agency

Answer: The Frontiersin.org review (published 18 August 2026) applied a four-phase framework (Questioning, Observing, Reasoning, Discovering) to classify epistemic agency in papers using the term "quantified self".

According to Frontiersin.org, the authors searched Web of Science and PubMed in April–May 2025 for the keyword "quantified self", yielding 721 unique records before screening.

According to Frontiersin.org, the authors retained 241 full papers for analysis and identified 46 studies where self-tracking was the data-collection method and could be assessed for agency.

The Frontiersin.org team categorized agency conservatively: if a phase could be attributed ambiguously they assigned the lower level, and disagreements were resolved by author discussion, per the Methods (Frontiersin.org, 18 August 2026).

The review quotes philosopher Catherine Elgin to define epistemic agency: "Epistemic agents should think of themselves as, and act as, legislating members of a realm of epistemic ends: they make the rules, devise the methods, and set the standards that bind them, " attributed to Catherine Elgin in the review (Frontiersin.org, 18 August 2026).

Findings Snapshot

Date / Source	Metric / Measure	Value (from review)	Implication for qualitative research
April–May 2025 search, reported in Frontiersin.org (18 Aug 2026)	Unique records retrieved	721 records	Corpus is broad; keyword choice shapes what literature is captured
Screening results, Frontiersin.org (18 Aug 2026)	Full papers retained	241 papers	A focused corpus for conceptual synthesis of QS in health
Agency-classified subset, Frontiersin.org (18 Aug 2026)	Papers using self-tracking methods	46 papers (17 low, 10 mixed, 19 high)	About 41% of self-tracking studies showed high participant agency
Publication year distribution, Frontiersin.org (18 Aug 2026)	Peak and recent counts	2018: 38 articles; 2021: 14; 2024: 13	Term usage peaked in 2018 and diversified in later years

Implications for qualitative researchers and research teams

Answer: The Frontiersin.org review (published 18 August 2026) implies qualitative teams must code agency explicitly and report who formulated the question when analyzing self-tracking data.

Qualitative researchers should annotate each case for the four phases (Questioning, Observing, Reasoning, Discovering) because the Frontiersin.org framework shows those phases map directly onto levels of epistemic control.

When a study treats participants only as data collectors, the review found it is classified as low agency; researchers should therefore distinguish contributions that are 'data supply' from those that are 'self-authored inquiry' (Frontiersin.org, 18 August 2026).

In practical terms, the review highlights that high-agency cases often name the self-tracker as co-author and that sharing of methods and learnings, not just raw data, is a marker of personal science (Frontiersin.org, 18 August 2026).

Ethics note: This synthesis is research-focused and not clinical advice; personal science projects may raise privacy and validity concerns that require ethical safeguards, as the review discusses (Frontiersin.org, 18 August 2026).

How Evidano Helps researchers apply the agency framework

Problem: Large, mixed corpora hide agency patterns

Answer: Researchers need reproducible thematic and cross-segment analyses to surface who authored questions and who performed reasoning.

Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.

Feature mapping: Use Evidano's thematic coding and hierarchical codes→subcodes to tag Questioning, Observing, Reasoning, and Discovering at scale, then run frequency and cross-segment analyses to report how many cases show high agency.

Contextual link: Learn about the platform capabilities on the Evidano features page.

Problem: Manual synthesis misses N-of-1 patterns

Answer: Personal science reports require integrating quantitative streams and qualitative reflections from the same person.

Feature mapping: Evidano ingests transcripts, CSV logs, and documents to produce combined thematic and frequency analyses so teams can identify instances where the person 'owned the question' and performed Reasoning and Discovering.

Practical note: Evidano visualizations (co-occurrence networks and hierarchical code maps) make it straightforward to show where subjective insights align with objective measurements.

Problem: Reconstructing methods and authorship for transparency

Answer: Transparent method-sharing is an indicator of high agency that qualitative reviewers should capture and publish.

Feature mapping: Use Evidano to extract method descriptions, code for method-sharing versus raw-data-sharing, and produce exportable audit trails for reproducibility and ethics review.

Security note: Evidano's data handling is designed for research contexts; see the Evidano data security page for details.

FAQ: epistemic agency in self-tracking

What is epistemic agency in self-tracking?

Answer: Epistemic agency in self-tracking means the individual both poses the research question and meaningfully participates in collecting, analyzing, and applying findings, as defined in the Frontiersin.org review (18 August 2026).

Supporting detail: The review uses the four-phase framework Questioning–Observing–Reasoning–Discovering to operationalize agency, and explicitly contrasts low-agency studies where participants only supply Observing with high-agency personal science where all phases are self-directed (Frontiersin.org, 18 August 2026).

How common is high-agency personal science in the published literature?

Answer: The Frontiersin.org review (18 August 2026) found 19 high-agency studies among 46 self-tracking studies it analyzed.

Supporting detail: In the review's classified subset of 46 studies, 17 were low agency, 10 mixed agency, and 19 high agency, showing substantial representation of personal science in that corpus (Frontiersin.org, 18 August 2026).

Can AI-enabled qualitative analysis distinguish who 'owns' the question?

Answer: Yes, AI-enabled qualitative analysis can reliably surface authorship signals and phase attribution when trained to code Questioning, Observing, Reasoning, and Discovering.

Supporting detail: The Frontiersin.org framework provides explicit coding targets; AI tools that ingest transcripts and documents can tag statements like 'I wanted to test whether' as Questioning and 'I analyzed my logs' as Reasoning, enabling reproducible counts and exemplars.

What are practical next steps for teams doing citizen science for health?

Answer: Teams should adopt transparent method-sharing, tag agency phases in their outputs, and consider infrastructure for ethical oversight as recommended by the Frontiersin.org review (18 August 2026).

Supporting detail: The review recommends education, community documentation, and ethical guidelines to support personal science without treating it as anecdote or mere data supply (Frontiersin.org, 18 August 2026).

Conclusion & Next Steps

Answer: The Frontiersin.org review (18 August 2026) shows that classifying epistemic agency clarifies whether self-tracking is measurement or personal science, and AI-enabled qualitative tools make that classification scalable and auditable.

Researchers can reproduce the review's approach by coding the four phases and counting agency levels, then publishing method extracts to show who owned each question (Frontiersin.org, 18 August 2026).

If you want to apply this approach to your corpus of interviews, self-experiments, or mixed-method case reports, try an AI-first qualitative workflow that extracts phases, authorship signals, and method-sharing.

Get started and Try Evidano for free.

Topics

epistemic agency in self-tracking
personal science qualitative analysis
quantified self agency
AI-enabled qualitative research

Keep reading

Browse all articles