Thematic analysis and interpretative phenomenological analysis Flashcards
approaches to qualitative data and analysis
- Grounded theory
- Content analysis
- Discourse analysis
- Conversational analysis
- Narrative analysis
- Thematic analysis
- Interpretative phenomenological analysis (IPA)
thematic analysis
- An umbrella term for a set of approaches that share a focus on identifying themes within data
- It minimally organises and describes your data in detail
- Thematic analysis underpins many other forms of qualitative analysis.
what is a theme?
- A theme is a pattern of meaning that captures:
- Something important about the material
- A shared implicit or underlying meaning
- Emphasis is on meaning, not prevalence
where do themes come from?
- Date-driven/inductive - coding and theme development are data-driven (bottom-up)
- Theory-driven/deductive - shaped by existing theoretical constructs, which provide the ‘lens’ to code and develop themes (top-down)
- Most likely it’s a combination of both
six phases of TA (Braun and Clarke, 2006)
- Familiarisation oneself with the data
- Generating codes
- Searching for themes
- Reviewing potential themes
- Defining and naming themes
- Producing the report
familiarisation oneself with the data
- We start by reading all transcripts and taking initial notes
- You may ask things like?:
- What sort of assumptions are being made?
- How are certain groups characterised?
- What ideas are being drawn on?
initial coding
· Two types of coding (Braun and Clarke, 2013, p. 206):
- Selective coding - identifying relevant material
- Complete coding - line by line
· Codes = basic units of meaning
- A piece of coded text varies from a few words to a multi-sentence chunk (Miles and Huberman, 1994)
managing the coding process - from low to high-tech
· Use highlighters, pens, post-it notes, or comments function in word
· Make use of CAQDAS - computer-assisted qualitative data analysis software e.g., Nvivo
reviewing themes
· Read all the collated extracts for each theme, and consider whether they appear to form a coherent pattern
· You decide:
- Some candidate themes are not really themes (e.g., if there are not enough data to support them, or the data are too diverse) - drop them
- Others might be merged into each other (e.g., two apparently separate themes might form one theme).
- Other themes need to be split into separate themes
searching for themes
· Process of clustering together similar codes - which belong together
· Organise your codes into initial themes - a code can be promoted to a theme
· Start to think about the relationship between themes - what is the overall story?
themes or topic summaries?
· In TA, themes are conceptualised as patterns of shared meaning underpinned by a central concept
· Multi-faceted - perhaps cutting across several topics - and telling a story about the data
· Different from topic summaries - buckets that collect together everything or the main things participants communicated about a topic - a shared topic.
defining and naming themes and writing up
· Write a definition - a short description for each theme
· Do not just paraphrase the extracts presented - identify what is interesting about them and why
· Writing (interpretation, commentary) is an integral part of analysis, not something that occurs just at the end (as it does with statistical analyses).
problems of thematic analysis
· “Themes emerged” - themes don’t passively emerge from the data
· Howitt (2013) - at its worst, the analyst ‘sees’ five or six themes and then just looks for examples:
- Implies that the themes are there without researcher input
- No justification or explanation was given for the themes
- No criteria - no effort
advantages of TA
· It can be used to address most types of qualitative research questions
- ‘How is race constructed in workplace diversity training?’
- ‘What do people think of women who play traditionally male sports?’
· It can be used to analyse most types of qualitative data:
- Interviews
- Newspaper materials
- Naturally occurring conversations
- Websites
· It is not tied to a particular theoretical framework
- TA is theoretically flexible
· Techniques have many features in common with other qualitative methods (IPA, Grounded Theory)
IPA
· A form of thematic analysis that makes a number of psychological assumptions
· IPA - interviews are for the study of experience (phenomenology)
(lived) experiences of what?
· Answer - people’s ‘life-worlds’
· The state of affairs in which the world is lived and experienced i.e., the subjective (not ‘behaviour’)
· What matters to participants
assumptions about knowledge in IPA
· 1st assumption:
- People interpret the world of phenomena (things)
- What we, researchers, study, therefore, are their interpretation (s) of their world
· 2nd assumption:
- Researchers interpret the world, too
- When we study people’s sense-making, we bring our own sense-making to this enterprise
- Researchers, interpret people’s interpretations
- Therefore, reflexivity (self-awareness in our activity as researchers) is built into IPA
you would do IPA if
· Idiographic vs nomothetic:
- You are interested in the particular cases, not general statistical (average) tendencies in the population (the notional ‘average person’ - who doesn’t actually exist)
· Meanings vs casual relations:
- You are interested in what things mean to people
- If you want to make claims about casual relations (e.g., the mechanisms in people’s heads that make them interpret in these ways) you should do experiments instead
· Quality vs quantity:
- You are interested in the quality or types of experiences, not in measuring amounts or strength of experiences (or anything else)
aims and research questions
· Experiences/events with major significance to the person
· People’s experiences and/or understandings of particular participants
· Perceptions and views of particular participants
· Research questions are open and exploratory and tend to focus on the process rather than the outcome:
- How does someone make sense of a major event/transition in their life?
- How does someone make an important decision?
data collection and samples in IPA
· Detailed examination of a particular case or a small number of cases:
- Small samples (usually no more than 10; 6-8 as standard)
- Typically, semi-structured interviews
- Diaries can be used too (not things like newspaper articles, etc)
sample size for an IPA interview study
· Sample type is more important than size (since the interest in the shared experience of a given situation or thing)
- Hence homogeneous rather than representative sample (idiographic not nomothetic)
analytical satges (Smith et al, 2009)
· Idiographic commitment:
- Understanding the participant’s point of view
- Psychological focus on personal meaning-making
· Set of common processes:
- Case by case - start with one, then move on to a second case, and so on
- From the particular to the shared/common
From description to interpretation
read through case 1
· IPA always begins with a detailed reading and analysis of a single case
· Read through transcript 1 (several times) - “Immersion”
- Insights come from knowing your data
identifying keywords or phrases
· Equivalent to ‘coding’ in thematic analysis:
- Put them in the margin or highlight them
· What are keywords?:
- The words that seem important as reflections of the speaker’s experience
- You have to make a judgement about this (remember the two assumptions about knowledge)
identify themes
· The ‘keywords’ you have highlighted may be indicative of possible themes (but there are not themes)
clustering themes
· Establish any connections between themes
· Clusters are superordinate themes:
- Cluster 1, psychological states - excitement, hunger for experiences
- Cluster 2, psychological changes - maturity, less risk-taking
· Not all themes may fit the clusters
· When you move on to analyse other cases, some of the themes might be dropped.
integration of cases
· Integration of cases - if you have more than one case
· Use the themes you have derived from the first interview as ‘hypotheses’ for the organisation of the second and third interviews
validation
· How can I be confident in the interpretations that make up my analysis?
· How can others have confidence in my interpretations?
· Why is one interpretation better than another?
· The validity of your analytic claims is ultimately a matter of their plausibility
validation (during the analysis)
· Iterative reading - stick to the data and return to the data:
- As you start to think about a possible theme, you ‘test’ it as you go back and forth from transcript to margin (‘Is this ‘excitement’? Is it like my other examples?)
· Ask yourself:
- Does this theme reflect an important and distinctive aspect of the speaker’s experience?
- Are these different instances similar - do they make up a theme (then merge)?
- Is this theme (‘excitement’) really different from the other theme (‘hunger for experience’) (then split)?
validation (in the write up)
· Present illustrative quotes for each of your themes
· Sufficient extracts from participants are presented to make a plausible case
· Then other people can judge for themselves whether your reading is plausible
