On the method-decomposing of reflective thematic analysis: for Housing Studies students
Highlight 5
main ideas of Virginia Braun and Victoria Clarke's work on reflective thematic
analysis and 3 main claims of the method in terms of Toulmin's model of
argument.
Based on Braun and
Clarke’s extensive writings on reflexive thematic analysis (RTA), here are five
central ideas of their approach, followed by three main claims of the method
framed using Toulmin’s model of argument.
Five main ideas in Braun and Clarke’s reflexive
thematic analysis
1.
Themes are
actively constructed, not “emergent”
Braun and Clarke insist that themes do not simply “emerge” from the data
waiting to be discovered. Instead, themes are interpretive outputs created by
the researcher through engagement with the data, their theoretical assumptions,
and their analytic skills.
2.
Researcher
subjectivity and reflexivity are analytic resources
Reflexive TA treats the researcher’s subjectivity, positionality, and
theoretical commitments not as biases to be eliminated but as essential
resources that shape and enrich knowledge production. Researchers are expected
to reflect critically on how their assumptions influence coding, theme
development, and interpretation.
3.
Theoretical
flexibility within coherent philosophical positioning
RTA is theoretically flexible and can be used within different epistemological
and ontological frameworks (e.g., essentialist vs constructionist; experiential
vs critical; inductive vs deductive; semantic vs latent). However, researchers
must be explicit and coherent about their philosophical stance and ensure it
aligns with their analytic choices and reporting.
4.
A recursive
six-phase analytic process
Braun and Clarke outline six overlapping, recursive phases: (1) familiarisation
with the data; (2) systematic coding; (3) generating initial themes; (4)
reviewing and refining themes; (5) defining and naming themes; and (6) writing
up the analysis. This process is iterative rather than strictly linear, with
constant movement between data, codes, and themes.
5.
Quality rests on
thoughtful engagement, not procedural correctness
High-quality reflexive TA is not about following a rigid checklist, achieving
inter-coder reliability, or producing “accurate” codes. Quality depends on the
researcher’s reflective and thoughtful engagement with the data and analytic
process, producing coherent, distinctive, and meaningful thematic
interpretations that answer the research question.
Three main claims of reflexive TA in
Toulmin’s model of argument
Toulmin’s model analyses
arguments in terms of claim (conclusion), data/grounds
(evidence), warrant (reasoning linking data to claim), plus backing,
qualifiers, and rebuttals. Framed this way, reflexive TA makes
the following core argumentative moves:
1.
Claim: Qualitative
knowledge is co-produced by researcher and data
o Data/grounds: Empirical materials (e.g., interview
transcripts) show multiple, context-dependent meanings.
o Warrant: Because meaning is socially and linguistically
constructed, any interpretation necessarily involves the researcher’s
theoretical and subjective lens.
o Backing: Constructionist and interpretivist traditions in
qualitative research; Braun and Clarke’s articulation of RTA as situated at the
intersection of dataset, theory, and researcher.
o Qualifier: This claim holds for interpretive qualitative
research, not for positivist approaches seeking single objective truths.
o Rebuttal addressed: Against views that researcher subjectivity
is “bias,” RTA reframes it as an analytic resource.
2.
Claim: Themes are
interpretive stories, not summaries of topics
o Data/grounds: Coded extracts reveal patterns of meaning
that can be organised around central organising concepts.
o Warrant: If themes are to capture meaningful patterns
relevant to the research question, they must be developed through active
interpretation, not just topic-based grouping.
o Backing: Braun and Clarke’s distinction between
“themes-as-meaning-unified-interpretative-stories” and
“themes-as-topic-summaries,” and their emphasis on latent and semantic coding.
o Qualifier: This applies when the aim is interpretive
understanding rather than descriptive cataloguing.
o Rebuttal addressed: Against coding reliability models that
treat themes as stable categories, RTA argues themes are creative analytic
outputs.
3.
Claim: Rigour in
TA comes from reflexive coherence, not procedural standardisation
o Data/grounds: Published TA studies often misuse TA by
mixing incompatible assumptions (e.g., claiming constructionism while seeking reliability
and “emergent” themes).
o Warrant: If methodological components (epistemology,
methods, analysis, reporting) are misaligned, the analytic argument becomes
incoherent and less credible.
o Backing: Braun and Clarke’s critiques of “positivist
drift” in TA and their call for “theoretical knowingness” and methodological
congruence.
o Qualifier: This rigour criterion is specific to reflexive,
interpretive TA, not to all forms of thematic or content analysis.
o Rebuttal addressed: Against demands for inter-coder agreement
as the gold standard, RTA argues that reflexive transparency and coherence are
more appropriate quality markers for this approach.
Together, these
claims position reflexive TA as an interpretive, researcher-centred method that
constructs thematic arguments through transparent, theoretically informed, and
reflexively managed analytic work.
Highlight
the main steps involved for this method and provide a brief example on its
practice reference to a Housing Studies research theme on the homelessness of
the poor elderly citizens in Hong Kong.
Main steps in Braun and Clarke’s reflexive
thematic analysis
Braun and Clarke
describe reflexive thematic analysis (RTA) as a six-phase, recursive process.
The phases are not a rigid checklist but overlapping activities that support
interpretive, theory-informed analysis.
1.
Familiarisation
with the data
o Read and re-read transcripts, field notes, or
documents.
o Take initial analytic notes on patterns,
surprises, and possible meanings.
o Begin thinking about how the data might
answer the research question.
2.
Generating initial
codes
o Systematically code the entire dataset, line
by line or segment by segment.
o Codes are interpretive labels that capture
something meaningful in relation to the research question (not just topic
tags).
o Keep memos on why you coded as you did and
what assumptions you are bringing.
3.
Constructing initial
themes
o Group related codes into broader patterns of
shared meaning (candidate themes).
o Use thematic maps, diagrams, or tables to
visualise relationships between codes and themes.
o Themes at this stage are provisional “working
ideas,” not final products.
4.
Reviewing and
developing themes
o Check each candidate theme against the coded
extracts and the full dataset.
o Ask: Does the theme have enough data? Is it
internally coherent? Does it tell a clear story?
o Merge, split, discard, or rename themes as
needed to improve coherence and relevance.
5.
Refining,
defining, and naming themes
o Write a detailed analytic description of each
theme: what it captures, its boundaries, and how it relates to the research
question.
o Choose names that reflect the interpretive
content of the theme, not just its topic.
o Identify sub-themes where appropriate to
capture nuance.
6.
Writing up the
analysis
o Weave together data extracts and your
interpretive narrative; quotes illustrate and support your argument, they do
not replace it.
o Situate your thematic story within existing
literature and your theoretical framework.
o Be transparent about your analytic decisions
and reflexive stance.
Brief worked example: Homelessness of poor
elderly citizens in Hong Kong (Housing Studies)
Research focus: Understanding how and why poor elderly
citizens in Hong Kong experience homelessness, and what this reveals about
housing policy, family support, and urban inequality.
1. Familiarisation
You collect and
read:
- 20 semi-structured
interviews with homeless elderly people in Hong Kong (e.g. sleeping in
parks, 24-hour fast-food outlets, or subdivided units).
- Field notes from outreach
workers and social workers.
- Policy documents on public
housing, CSSA (Comprehensive Social Security Assistance), and elderly
support services.
While reading, you
note recurring ideas such as “waiting too long for public housing,” “family
conflict,” “health problems,” and “stigma about asking for help.”
2. Initial coding
From one interview
excerpt:
“I applied for
public housing five years ago, but they said my son owns a flat, so I’m not
eligible. But my son doesn’t let me live with him. I don’t want to burden him…
I sleep at the McDonald’s at night.”
Possible codes:
- “Long public housing waiting
time”
- “Eligibility rules tied to
family assets”
- “Strained intergenerational
relations”
- “Reluctance to burden
family”
- “Use of 24-hour commercial
spaces as shelter”
You apply similar
interpretive codes across all interviews and documents.
3. Constructing initial themes
Grouping codes,
you develop candidate themes such as:
- “Policy
exclusion through family-based eligibility” (codes about asset tests, children’s property, waiting lists).
- “Filial
piety, shame, and silence” (codes about not wanting to
burden family, hiding homelessness, emotional distress).
- “Precarious
survival spaces in the city” (codes about sleeping in
fast-food outlets, parks, subdivided units, fear of police).
4. Reviewing themes
You check each
theme:
- Does “Policy exclusion”
capture enough data across cases? Yes – many mention asset rules and long
waits.
- Is “Filial piety, shame, and
silence” coherent? You notice some elderly are estranged and no longer
care about shame; you refine the theme to focus on “Moral expectations of
family care and their limits.”
- You consider whether
“Precarious survival spaces” should be split into “Informal shelter
strategies” and “Risk and insecurity in public/private spaces.”
5. Refining, defining, and naming themes
Final themes might
be:
1.
“Structural
exclusion from formal housing support”
Captures how eligibility rules, long waiting times, and asset tests
systematically exclude poor elderly from public housing, even when family
support is absent in practice.
2.
“Negotiating
filial obligations and dignity”
Shows how elderly people balance cultural expectations of family care with
realities of conflict, abandonment, or desire not to be a burden, often leading
to silence about their homelessness.
3.
“Living on the
edge: informal and risky shelter practices”
Describes how elderly homeless people rely on 24-hour commercial spaces, public
areas, and overcrowded subdivided units, facing constant insecurity, health
risks, and policing.
Each theme is
defined in a paragraph, with sub-points and clear links to the research
question about causes and experiences of elderly homelessness.
6. Writing up
In your Housing
Studies dissertation or article, you:
- Introduce each theme with an
analytic narrative.
- Use short, carefully chosen
extracts (e.g. the quote above) to illustrate and support your
interpretation.
- Connect your themes to
literature on Hong Kong’s housing policy, filial piety, welfare regimes,
and urban homelessness.
- Reflect on your
positionality (e.g. as a researcher, possibly middle-class, speaking with
very poor elderly participants) and how that shaped your questions and
interpretation.
This example shows
how RTA moves from raw interview and policy data to an interpretive,
theory-informed thematic story about elderly homelessness in Hong Kong,
suitable for a Housing Studies research project.
A collection of blog notes on using chatgpt for research purpose.
No comments:
Post a Comment