The writing is where an interpretation becomes testable

This tenth of ten units argues that writing is part of the analysis rather than a report on it, so the record of decisions is evidence and the form of the account carries a claim. The nine units before it each settled a question of evidence.

The writing is a method of inquiry, not a report

The writing is not a report of work that finished earlier. The record of decisions names which readings were taken, which were resisted, and what would have to be true for a rival reading to win.

Richardson (2000) made the founding move, arguing that writing is a method of inquiry in which the act of writing produces understanding rather than merely reporting it. Braun and Clarke (2021) place the writing of the report inside the analysis: in reflexive thematic analysis the analyst's interpretive work is part of the method, not a bias to be subtracted. Charmaz (2014) reaches the same place through memo-writing, which she treats as a core analytic practice in constructivist grounded theory.

Reflexivity is documentary before it is confessional

Finlay (2002) defines the practice. Reflexivity is explicit, self-aware analysis of the researcher's own role, undertaken because intersubjective elements shape what is collected and what is made of it. She argues that this evaluation secures the study's trustworthiness. She is candid that the process is difficult and contested, so that researchers may retreat from it. A documentary standard answers that retreat: a record of the decisions is the alternative to omitting them.

Mauthner and Doucet (2003) name the distinction a documentary standard has to keep. Writing from their own doctoral analysis, they separate two things. Accounts of reflexivity narrate the researcher's relation to the study. Reflexive accounts expose the analytic decisions that produced a reading. The difficulties of reflexivity are rarely addressed, and a retrospective account can tidy messy interpretive work into a story with no decisions in it.

Mauthner studied motherhood and postnatal depression, and Doucet studied couples sharing childcare and paid work. Doucet reports that the birth of her three children during the five years of the doctorate changed how she read her interviews about parenting and employment. That influence can be told as a story about the researcher, or it can be recorded as an interpretive decision, and only the second is evidence about the analysis.

Reflexivity is asked for and practised unevenly

Olmos-Vega and colleagues (2023) give the practice its current form in AMEE Guide No. 149. They define reflexivity as continuous, collaborative and multifaceted practices through which researchers appraise how their subjectivity and context influence the research. They separate the personal, interpersonal, methodological and contextual factors that bring a study into being. Their demand that reflexivity be communicated as part of the method, not added retrospectively, states the documentary standard as a design requirement.

Probst (2015) supplies the empirical check. She interviewed 34 qualitative social work researchers from eight countries, 65 per cent in the United States, about how they practised reflexivity. They described benefits for themselves and their projects, and obstacles that were personal, project-related, professional and systemic. They described a discrepancy between how much they valued reflexivity and how much they used it.

Position shapes what is noticed, and position moves

Berger (2015) shows why the record cannot be a single paragraph. She examines three types of researcher position, distinguished by how far the researcher shares the participants' experience. The same setting yields different observations as position changes, so a claim about what the data show depends on where the analyst stood.

Her own three cases show the range. She studied immigrant women's experience a decade after their move to a new country. She studied stepfamilies before and while she became part of one, by marrying a divorced father who took the stepfather role for her son. She studied women who had been abused and divorced, a situation she had not shared.

Moralli (2024) takes the point further in migration research in Italian shrinking areas, and proposes research as care. Her account turns on the epistemological importance of emotions, intersectional positionality and an ethics of care. She describes research as a negotiation in which positionality is flexible and conditioned by the emotionalities that emerge in the field.

The form of the account carries a claim

Van Maanen (2011) shows that the narrative form of an account is itself a claim. He distinguishes three forms. Realist tales report a settled field in the researcher's voice. Confessional tales foreground the researcher's experience. Impressionist tales evoke the field through striking scenes. The choice among them is methodological, because each sets the reader's trust on a different footing.

Blandford, Furniss and Makri (2016) bring the argument to human-computer interaction. Much qualitative guidance there comes from the social sciences rather than from research on systems. Following a study from planning to reporting, they set out two disciplines for a findings section. Separate the participant's words from the analyst's claim. Report the disconfirming evidence, naming the case that does not fit rather than dropping it. When the interview was conducted in a second language, that separation is also a translation, and the account should say who made it.

Paulus and colleagues (2017) show how sparsely the record is kept. Their discourse analysis covered 763 empirical articles published between 1994 and 2013. Most authors named the program but gave few details of how they used it. Their recommendations are documentary. Choose active voice so that the software is not credited with the analysis, define specialist terminology, and avoid the unsubstantiated claim that using the software improved the quality of the work.

Contested question: is the reflexive record evidence, or its performance?

The disagreement is not whether analysts should be reflexive. It is whether the written record is evidence of an inspected analysis, or a performance that leaves the analysis untouched. The literature holds three positions.

• Documentary. The record exposes the decisions and can be assessed: reflexivity is explicit self-aware analysis of the researcher's role (Finlay, 2002); a planned practice communicated in dissemination (Olmos-Vega et al., 2023); and something whose adequacy a study can be asked to evidence (Probst, 2015).
• Performative. Reflexivity is a methodological power that reassures the reader without disturbing the researcher's authority; Pillow (2003) identifies four comfortable uses, as recognition of self, of other, as truth and as transcendence, against which she sets uncomfortable practice.
• Concealing. Mauthner and Doucet (2003) hold that a reflexive account can be a tidy narrative that hides the analytic work, so a record can be present and still not be evidence.

The evidence that would separate them has not been gathered. Write two versions of one study's account, the second adding the decision record: the codes tried, the extract that resisted the reading, and the evidence that would defeat it. Give them to reviewers blind to the version. Ask both whether the interpretation is warranted by the evidence, and whether a second analyst could reconstruct the coding decisions. If the record moves the first judgement, it is evidence. If it moves only the second, or only a rating of thoroughness, Pillow's reading holds and the record is a display. A reproduction test is sharper. If a second analyst working from the record arrives at different themes, the account did not carry the analysis. The dispute reaches coding itself: Braun and Clarke (2019) distinguish reflexive thematic analysis from codebook and coding-reliability approaches, arguing that treating themes as waiting to be discovered misstates the method.

The literature is rich in examples and poor in tests of its own claim. Probst (2015) interviewed 34 researchers about their practice, and Paulus and colleagues (2017) analysed 763 articles to show how sparsely the record is kept. Both measure what is written, not whether it is evidence. Pillow (2003) and Mauthner and Doucet (2003) argue the performative and concealing cases from their own work. While the dispute is open, a designer should keep the decision record as the analysis proceeds, state which readings were resisted and what would defeat the chosen one, and write so that a second analyst could reconstruct the coding. The strongest objection is that a decision record is retrospective by construction, so it can be written after the reading has settled and still read as though it guided it.

Where the argument is heading

The direction of travel is from reflexivity placed before the findings to reflexivity woven through the analysis and the writing. Moralli (2024) puts positionality at the centre as something negotiated in the field. Olmos-Vega and colleagues (2023) ask for reflexivity aligned to a declared paradigmatic stance. Braun and Clarke (2021) keep the writing inside the analysis.

The sources this unit rests on

Each note states what the cited document did and found and what it means for the argument. Records were checked against Crossref and OpenAlex on 31 October 2026.

• Richardson, L. (2000). Writing: a method of inquiry. In N. K. Denzin and Y. S. Lincoln (eds), Handbook of Qualitative Research, 2nd ed. Thousand Oaks, CA: Sage. https://methods.sagepub.com/book/handbook-of-qualitative-research-2e Argues that writing is a method of inquiry in which the act of writing produces understanding rather than only reporting it, and supplies the founding claim that the report is part of the analysis. The note rests on the chapter record because the piece carries no DOI.
• Braun, V. and Clarke, V. (2021). Thematic Analysis: A Practical Guide. London: Sage. https://www.sagepub.com/shop/buy-a-book/thematic-analysis-1-248481 Gives the fullest statement of reflexive thematic analysis, from coding to developing themes and writing the report, then theory, interpretation and quality. It treats the analyst's interpretive work as part of the method and places the writing of the report inside the analysis.
• Charmaz, K. (2014). Constructing Grounded Theory, 2nd ed. London: Sage. https://archive.org/details/constructinggrou0000char Sets out constructivist grounded theory, covering coding, memo-writing and theoretical sampling. It treats memo-writing and reflexivity as core practices and argues that the researcher constructs the theory rather than discovering it.
• Finlay, L. (2002). "Outing" the researcher: the provenance, process, and practice of reflexivity. Qualitative Health Research, 12(4), 531-545. doi:10.1177/104973202129120052 Argues that qualitative research secures integrity and trustworthiness by evaluating how intersubjective elements shape the work, and defines reflexivity as explicit, self-aware analysis of the researcher's role. It reports that the process is difficult and contested, so researchers may retreat from it.
• Mauthner, N. S. and Doucet, A. (2003). Reflexive accounts and accounts of reflexivity in qualitative data analysis. Sociology, 37(3), 413-431. doi:10.1177/00380385030373002 Draws on the authors' doctoral analysis to separate accounts of reflexivity, which narrate the researcher's relation to the study, from reflexive accounts, which expose the analytic decisions. It argues that a tidy reflexive narrative can hide the interpretive work.
• Olmos-Vega, F. M., Stalmeijer, R. E., Varpio, L. and Kahlke, R. (2023). A practical guide to reflexivity in qualitative research: AMEE Guide No. 149. Medical Teacher, 45(3), 241-251. doi:10.1080/0142159x.2022.2057287 Defines reflexivity as continuous, collaborative and multifaceted practices through which researchers appraise how their subjectivity and context influence the research, and separates personal, interpersonal, methodological and contextual factors. It argues reflexivity belongs at all stages and is communicated as part of the method (2022 online-first; 2023 issue).
• Probst, B. (2015). The eye regards itself: benefits and challenges of reflexivity in qualitative social work research. Social Work Research, 39(1), 37-48. doi:10.1093/swr/svu028 Reports an empirical study of how reflexivity is practised, interviewing 34 qualitative social work researchers from eight countries, 65 per cent in the United States. It finds a discrepancy between valuing reflexivity and using it, names obstacles across personal, project-related, professional and systemic lines, and asks how a study's reflexivity might be assessed.
• Berger, R. (2015). Now I see it, now I don't: researcher's position and reflexivity in qualitative research. Qualitative Research, 15(2), 219-234. doi:10.1177/1468794112468475 Examines three types of researcher position, distinguished by how far the researcher shares the experience of participants, and illustrates the benefits and challenges of each with case examples. It argues that position is not fixed, so the same data can be read differently as it shifts (2013 online-first; 2015 issue).
• Moralli, M. (2024). Research as care: positionality and reflexivity in qualitative migration research. Qualitative Research, 24(3), 751-770. doi:10.1177/14687941231176946 Analyses reflexivity and positionality in migration research in Italian shrinking areas and proposes research as care, organised around the epistemological importance of emotions, intersectional positionality and an ethics of care. It shows that positionality is flexible and conditioned by the emotionalities of fieldwork (2023 online-first; 2024 issue).
• Van Maanen, J. (2011). Tales of the Field: On Writing Ethnography, 2nd ed. Chicago: University of Chicago Press. https://archive.org/details/talesoffieldonwr00vanmrich Distinguishes realist, confessional and impressionist tales in ethnographic writing and shows how each form shapes what the account claims about the field and where it places the reader's trust.
• Blandford, A., Furniss, D. and Makri, S. (2016). Qualitative HCI Research. Cham: Springer. doi:10.1007/978-3-031-02217-3 Addresses the gap in HCI-specific guidance on qualitative methods, following a study from planning to reporting, and argues that qualitative methods matter in HCI for requirements, understanding use and evaluating practice. It sets out two disciplines for a findings section: separating the participant's words from the analyst's claim, and reporting disconfirming evidence.
• Paulus, T., Woods, M., Atkins, D. P. and Macklin, R. (2017). The discourse of QDAS: reporting practices of ATLAS.ti and NVivo users with implications for best practices. International Journal of Social Research Methodology, 20(1), 35-47. doi:10.1080/13645579.2015.1102454 Analysed the language of 763 empirical articles published between 1994 and 2013 to see how researchers report the use of qualitative data analysis software, and found most gave few details beyond naming the program. It recommends active voice so the software is not credited with the analysis (2015 online-first; 2017 issue).
• Pillow, W. (2003). Confession, catharsis, or cure? Rethinking the uses of reflexivity as methodological power in qualitative research. International Journal of Qualitative Studies in Education, 16(2), 175-196. doi:10.1080/0951839032000060635 Identifies four comfortable uses of reflexivity, as recognition of self, of other, truth and transcendence, and argues that comfortable reflexivity can legitimise research without disturbing the researcher's authority. It calls for uncomfortable reflexive practices and reviews three authors who practise them.
• Braun, V. and Clarke, V. (2019). Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health, 11(4), 589-597. doi:10.1080/2159676x.2019.1628806 Reviews the assumptions behind their 2006 paper on thematic analysis and revises them, preferring the name reflexive thematic analysis and distinguishing it from codebook and coding-reliability approaches. It argues that treating themes as waiting to be discovered and coding as reliability-checkable misstates the method.