Epistemic Agency in Self-Tracking: AI for Qualitative Research

Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents. Qualitative researchers and UX teams wrestling with self-tracking data face a specific problem: how to tell whether a study treats the person as a passive data source or as the author of their own inquiry. The primary keyword for this post is "epistemic agency in self-tracking" and this article explains practical, extractable methods from a conceptual literature review that classified 721 records into a 241-paper corpus and then analyzed 46 self-tracking studies, according to the Frontiersin.org review by Sara Riggare published 18 August 2026. The payoff: concrete coding prompts and AI workflows you can apply to transcripts, logs, and device exports to measure agency across Questioning, Observing, Reasoning, and Discovering.

Key Takeaways

According to the Frontiersin.org review by Sara Riggare (published 18 August 2026), an agency-based, four-phase framework (Questioning, Observing, Reasoning, Discovering) distinguishes low, mixed, and high epistemic agency in self-tracking; the review found 46 self-tracking studies within a 241-paper corpus drawn from 721 search hits in April–May 2025 (Frontiersin.org).

A Web of Science and PubMed search in April–May 2025 returned 721 records, 241 full papers were retained for analysis, and 46 papers used self-tracking as the data-collection method, according to Sara Riggare et al., Frontiersin.org, 18 August 2026.
The 46 self-tracking papers split into 17 low-agency, 10 mixed-agency, and 19 high-agency studies, as classified by the review (Riggare et al., published 18 August 2026).
The review identifies the decisive distinction as “who owns the question”; Riggare et al. write that when the individual defines the question and interprets results, self-tracking becomes personal science (Riggare et al., 18 Aug 2026).
A useful short formal quote to carry into codebooks is Catherine Elgin’s definition cited in the review: "Epistemic agents should think of themselves as, and act as, legislating members of a realm of epistemic ends" (Elgin, quoted in Riggare et al., 18 Aug 2026).

What Happened: How the review measured epistemic agency

Answer: The review operationalized epistemic agency as who performs each phase of inquiry and then coded articles accordingly. According to the Frontiersin.org article by Sara Riggare (published 18 August 2026), the authors adapted a personal science cycle into four phases (Questioning, Observing, Reasoning, Discovering) and applied three agency levels: low, mixed, high.

According to Riggare et al. (18 Aug 2026), the literature search used the term "quantified self" in Web of Science and PubMed in April–May 2025, yielding 721 unique records; the authors excluded non-empirical pieces and retained 241 full papers for reading. According to the review, of those 241 papers, 46 used self-tracking as the data collection method and were analyzed for agency.

According to the Frontiersin.org review (Riggare et al., 18 Aug 2026), low-agency articles (n = 17) treated participants as data sources where only Observing was attributed to the person; mixed-agency articles (n = 10) showed partial participation; and high-agency articles (n = 19) showed the full self-directed cycle and often listed the self-tracker as a co-author.

Method constraints noted by the authors included: conservative coding rules that defaulted to the lower agency level in ambiguous cases, dual-reviewer arbitration (13% abstract conflicts resolved in April–May 2025), and the authors' positionality as insiders to the Quantified Self community (Riggare et al., 18 Aug 2026).

Findings Snapshot

Date	Metric	Value	Implication
April–May 2025	Search hits for "quantified self"	721	Broad corpus; term is established but heterogeneous (Frontiersin.org, Riggare et al., 18 Aug 2026).
April–May 2025	Full papers retained	241	Filtered corpus where agency could be meaningfully assessed (Riggare et al., 18 Aug 2026).
April–May 2025	Papers using self-tracking as data collection	46	Subset directly analyzed for epistemic agency (Riggare et al., 18 Aug 2026).
18 August 2026	Agency breakdown among the 46	17 low / 10 mixed / 19 high	Shows the full spectrum from sensor-platform studies to personal science (Riggare et al., 18 Aug 2026).
2018	Publication peak using term	38 articles	Historical peak noted; field shifted terms and framing after 2018 (Riggare et al., 18 Aug 2026).

Implications for qualitative researchers studying self-tracking

Answer: Use an agency-first codebook that asks, for each case, who set the question, who designed the observation, who did the analysis, and who articulated the discovery. According to Riggare et al. (Frontiersin.org, 18 Aug 2026), the key analytic move is to code each of the four phases (Questioning, Observing, Reasoning, Discovering) by actor (individual, researcher, mixed).

Practical decision 1: When you plan interviews about self-tracking, include explicit probes about ownership of the research question and whether the participant iterated measurement methods, because the review found that high-agency cases often involved self-defined questions refined across data collection (Riggare et al., 18 Aug 2026).

Practical decision 2: When building qualitative codebooks, include nodes for "method sharing" versus "data sharing" because the review documents that low-agency studies foreground data sharing while higher-agency studies foreground sharing of methods and learnings (Riggare et al., 18 Aug 2026).

Practical decision 3: Treat N-of-1 reasoning artifacts (spreadsheets, visualizations, export logs) as primary qualitative sources, because the review reports that high-agency personal science commonly uses visualizations and iterative hypothesis testing as part of Reasoning (Riggare et al., 18 Aug 2026).

How Evidano Helps: map problems to AI-enabled features

Problem: Slow synthesis of mixed-format personal science artifacts

Solution: Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents. Evidano ingests interview transcripts, spreadsheet exports, and device logs and applies thematic extraction to surface which of the four agency phases (Questioning, Observing, Reasoning, Discovering) the participant performed.

Practical benefit: Automatically tag transcripts and uploads for phrases that indicate ownership (for example, "I wanted to know", "I wrote code to export"), then quantify frequency by participant and study to replicate the 17/10/19 split logic used in the Frontiersin.org review (Riggare et al., 18 Aug 2026). See Evidano features: Evidano features.

Problem: Time-consuming transcription and messy device exports

Solution: Evidano’s speech-to-text with PII redaction and custom dictionaries handles interviews with technical terms or branded device names common in self-tracking studies, reducing manual cleanup time.

Practical benefit: Faster turnaround for Reasoning-phase evidence extraction when participants describe visualizations or custom scripts; see Evidano speech-to-text.

Problem: Extracting cross-case patterns about "who owns the question"

Solution: Evidano’s cross-segment analysis and AI chat over documents let you ask targeted queries like, "Which participants authored their question? " and retrieve coded excerpts and counts, mirroring the analytical decisions used by Riggare et al. in the Frontiersin.org review (18 Aug 2026).

Practical benefit: Turn 46 heterogeneous cases into a replicable matrix of agency indicators, then export tables and visualizations for papers or stakeholder reports.

FAQ: epistemic agency in self-tracking

What is epistemic agency in self-tracking?

Answer: Epistemic agency in self-tracking is the capacity of an individual to pose questions, collect and analyze data, and state findings about themselves, rather than merely supplying measurements to external researchers. The Frontiersin.org review defines it as the individual acting as producer of knowledge (Riggare et al., 18 Aug 2026).

Supporting detail: The review links the concept to Miranda Fricker’s epistemic injustice literature and to Catherine Elgin’s definition that epistemic agents act as "legislating members" of epistemic practice (Elgin quoted in Riggare et al., 18 Aug 2026).

How did Riggare et al. measure agency in the literature?

Answer: They coded who performed each of four inquiry phases (Questioning, Observing, Reasoning, Discovering) across articles and assigned low, mixed, or high agency by conservative dual-author review. Riggare et al. describe this method in the Frontiersin.org paper published 18 Aug 2026.

Supporting detail: The authors started from 721 records found in April–May 2025 and analyzed 241 full papers, of which 46 contained self-tracking data (Riggare et al., 18 Aug 2026).

Can personal science findings be used in qualitative research?

Answer: Yes, personal science artifacts are legitimate qualitative data if treated as situated, reflexive evidence that combines subjective meaning with objective measures. The review positions personal science as a distinct empirical practice with goals of actionable self-knowledge (Riggare et al., 18 Aug 2026).

Supporting detail: Riggare et al. note that high-agency cases often integrate qualitative narrative and quantitative logs and sometimes publish as peer-reviewed case reports.

How can AI accelerate assessing epistemic agency?

Answer: AI can accelerate assessment by extracting ownership language, mapping actors to the four phases, and counting evidence items per case, turning qualitative patterns into reproducible metrics. The Frontiersin.org review shows that counting phase authorship (who did Questioning, Reasoning, etc.) is a tractable analytic move (Riggare et al., 18 Aug 2026).

Supporting detail: Use automated transcription, entity extraction, and thematic coding to flag phrases like "I designed" or "the researchers did", then aggregate frequencies across participants to mirror the review’s classification approach.

Conclusion & Next Steps

Answer: Measuring "epistemic agency in self-tracking" requires a phase-by-phase codebook and the ability to synthesize heterogeneous artifacts; the Frontiersin.org review by Sara Riggare (18 Aug 2026) provides a ready framework and empirical benchmark (17 low, 10 mixed, 19 high among 46 self-tracking papers).

Recap: Qualitative teams should adopt explicit probes for ownership of the question, preserve method-sharing artifacts, and quantify evidence of agency across Questioning, Observing, Reasoning, and Discovering when possible (Riggare et al., 18 Aug 2026).

Next steps: If you want to operationalize these steps on your dataset, use AI-assisted ingestion for transcripts and device logs, run thematic and cross-segment analyses, and export agency metrics for reporting. Try a hands-on trial: Try Evidano for free.

Topics

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

Keep reading

Browse all articles