Research methods and statistics 4 (year two) Flashcards
give the types of thematic analysis
- Thematic analysis
- IPA
- Grounded theory
- Narrative analysis, conversation analysis
describe what thematic analysis is
What?
• Searching through data to identify any recurrent patterns.
• A theme is a cluster of linked categories conveying similar meanings and usually emerge through the inductive analytic process which characterises the qualitative paradigm
When?
• TA is a method that is
“essentially independent of theory and epistemology and can be applied across a range of theoretical and epistemological approaches” (Braun & Clarke, 2006).
• Different forms of analysis are not mutually exclusive – you can combine different approaches to fit your research question.
Give Braun and Clarke’s stages of thematic analysis
- Transcribe and immerse yourself in the data - familiarise
- Develop initial codes - generate
- Searching for themes - organise
- Review themes – re-read, check and amend
- Define and name themes – finalise
- Write the report
describe the familiarism stage
- Start by reading your data a number of times (at least two!)
- Active reading – make notes on your initial thoughts, what is interesting in the data, are there any repetitions?
- Keep notes as these will act as the foundation for the next stage of analysis
describe the coding stage
- Work your way through the data in a systematic way - code areas that come up a lot or common language used
- Apply label (code using key word/or phrase - you can do this by hand or using word/NVivo - Code areas that come up a lot or common language used and look for important parts of the transcripts which look at particular topics
- Review coding as you go
- You can do this on a selection of transcripts to generate a coding framework (20-30%)
• You can code a segment of text to multiple codes
• Code generously in relation to your research question
• CODE INCLUSIVELY – remember to keep some surrounding data for context
• MEMOS – build on your initial note-taking and start to write memos to document and increase the transparency of your analysis.
describe the creating themes stage of thematic analysis
- Review and consolidate - Review and consolidate codes into broader categories (i.e., themes)
• Are there any duplicates?
• Can any codes be amalgamated into higher order codes (emotion example from video) - Generate - Generate broader categories and sub-categories (themes) - use mind maps, post-it notes etc
- Keep writing memos
describee the reviewing themes stage of thematic analysis
- Re-read individual themes and allocated data - Go back to the data and check that all extracts fit into that theme
- If extracts don’t fit into the theme? - Go back to the stage before stage 4, re-think, and re-categorise as necessary (you don’t want a red sock in a theme called ‘white washing’ – it will turn everything pink and skew the theme’s meaning)
- Review themes in relation to data set. - Do they reflect the meaning as a whole? You don’t want themes about types of washing if most of the dataset is about football!)
- If themes don’t reflect the meaning as a whole? - Go back the previous stage, re-think and amend the themes
describe the defining themes and write up stage of thematic analysis
- When your thematic framework is finished, critically consider what each theme is really about and decide on a final name for themes - Remember to interpret and explain and not just describe!
- Write a detailed analysis of each theme -
- What is happening in the data in relation to this theme?
- How can you explain this?
- Use sub-themes to give your analysis structure and clarity
what is grounded theory?
- Most widely used way of analysing qualitative data
- Offers a flexible set of inductive guidelines for collecting and analysing qualitative data
- Involves creating codes by defining what you see in the data, not according to preconceived ideas – codes “grounded” in the data
- Involves forming a theory based on the collected data as opposed to collecting data after forming a theory (i.e. as in quant research).
- Ultimate aim of grounded theory is to generate a theory appropriate to the data and justifiable by close examination of the data
- Lots of variation depending on theoretical stance: Glaser & Straus (1967), Strauss (1987), Charmaz (2003).
give some of the key characteristics of grounded theory
- Systematic – The process by which theory is generated is through the careful application of the general principles and methods of grounded theory
- Guidelines: GT is essentially a set of guidelines which guide data collection, analysis, and theory generation
- Inductive NOT deductive – focus on generating theories rather than confirming them
describe the origins of grounded theory
• The discovery of grounded theory (Glaser & Strauss, 1967)
• Closing the gap between theory and empirical research
• To provide a sound basis for thorough qualitative research
• They eventually split in the 1990s and developed divergent schools of thought
“Glaserian” and “Straussian” versions of grounded theory
• Then Charmaz too (but we wont delve that far)
• Consequence: confusion around exactly what grounded theory procedures and terms are in the literature!
• We use “Straussian” grounded theory in the Department of Psychology
Describe the differences between Glaser vs Strauss
Glaser • Starts with a completely empty mind • Has a less structured coding process • Researcher is passive • Anything can be used as data – even quant! Strauss • Has a general idea of where to start • Has a more structured coding process • Researcher is active • Only traditionally qualitative data
describe when to use grounded theory
- When you want to generate a theory
- Understanding of social processes, experiences of a phenomenon
- When there is little or no existing research available on a topic
- Exploratory, generates ideas for further research
- Can be used with any type of qualitative data: media content, conversations, focus groups, biographical data etc..
- Particularly well suited to interview data
- Need to decide on a grounded theory analysis before data collection starts due to the sampling methods involved
Give some of the key features of grounded theory
- Theoretical sensitivity
- Theoretical sampling
- Constant comparison
- Data saturation
- Memo-writing
- Stages of coding
describe theoretical sensitivity
• Step 1: enter the research with as few predetermined ideas as possible
• Literature review not carried out until after analysis
• Prevents extensive reading
• Researcher not totally naïve – already has ‘sensitising concepts’
• Step 2: The ability to generate concepts from data and relate them meaningfully
• Immersion: See from the respondents point of view
• Questioning: Who, when, what, where, how much, why?
• Researcher needs to be aware of implicit meanings of the data
• Theoretical sensitivity is creative, insightful, conceptual
•
Describe theoretical sampling
- Unique to grounded theory
- Also known as purposive sampling
- The process of selecting “incidents, slices of life, time periods, or people on the basis of their potential manifestation or representation of important theoretical constructs” (Patton, 2001, p. 238).
- Simultaneous data collection & analysis
- Start with an initial (a priori) sample
- Generally quite broad inclusion criteria
- Further cases included on the basis of their usefulness in generating theory
Describe and explain data saturation
- Theoretical saturation of data means that researchers reach a point in their analysis of data that theoretical sampling of more data will not lead to new information related to their research questions.
- No additional data can be found to generate new relationships and concepts
- Researchers see the same codes, categories, and themes in their data over and over again
- Empirically confident that data is saturated
- Guest, Bunce & Johnson (2012) – “Twelve interviews is enough”
Describe and explain memo writing
- Cannot be over-emphasised!
- Writing memos throughout analysis:
- Reflective memos during data collection
- Guide theoretical sampling
- Coding & theory generation
- Start as soon as you have interesting ideas
- Object is to increase theoretical sensitivity
- They evolve – need to come back to them as analysis progresses
- Should be conceptual and not descriptive
- Strauss & Corbin (1998):
- Vary in length & detail
- At the start ‘awkward and simple’
- Individual styles
- Force you to work with concepts, not data
- Force reflection & coherence
- Necessary at end of any analytic session
- Can be written from other memos, not just raw data
Give the stages of coding
- Open coding (codes)
- Axial coding (categories)
- Selective coding (themes)
- Theory formation
- Literature review
Describe and explain open coding
- Line-by-line (per transcript)
- Name each line of data
- Varied and numerous
- Quick, generative – can produce 100s of codes
- Most lines in the transcript should generate an open code but if there are a few blank that’s ok
- Each code should be one or two words (no more than three)
- If more than 1 words, the words should be linked (usually an adjective and noun)
- Try to use words which the respondent uses.
- Gives you leads to pursue
- Start memo writing at this stage; open codes….
Describe and explain axial coding
- Creation of higher-order codes (“conceptual categories”) from open codes (per transcript)
- Only those that reflect data most accurately.
- Looking for:
- Relationships between categories
- Properties of categories
- Action-oriented
- Form categories by:
- Bunch frequent, similar codes into a single category.
- Can form a category based on 2 frequent, opposing codes (e.g. safe/fearful or anxious/relieved)
- Still memo-writing
- Look for quotes to support your categories (in the appendix)
describe and explain selective coding
- “The process of integrating and refining categories” (Strauss & Corbin)
- Now looking across all transcripts
- Form 4-5 themes based on axial categories
- All themes should be made up of categories which appear in at least two of the interviews.
- Bunch frequent, similar categories into a single theme
- Only those that reflect data most accurately.
- Now a lot more abstract & generalised
- Look for quotes to support these themes (to go in your results)
- Reflexive: Return to open and axial codes
- Start of theory formation
Describe the process of selecting a core theme
- Characteristics of a core theme:
- It’s central!
- Frequent appearance in data
- Abstract
- Can explain variation in the data
- How to choose the core theme?
- Reviewing earlier stages of coding
- Reviewing memos
- The focal part of an emerging storyline
- You generally just know by this point.
- But avoid circularity!
Describe the process of generating a core theme
- Generating relationships between the themes and both the core theme and each other - interrelationships
- Writing a storyline
- Moving from description to conceptualisation
- Diagramming (very common, my fave, goes in your appendix)
- Reviewing coding and memos
- Essentially, the product of this is the grounded theory
- Structure of the theory should emerge naturally from the themes in the analysis
- Does it answer the initial research question/s?
- Theory must be explanatory, not descriptive…
- No two grounded theories will look the same
Describe the process of selecting codes
- Let the respondents voice be heard
- Select either “typical” or “poignant” quotes
- Typical quotes highlight the essence of a theme or inter-relationship
- “e.g. theme of support – “she helps me emotionally and practically”
- Poignant quotes highlight the extent of the theme or inter-relationship – should be emotive
- E.g. “without her help I would have died”
- Quotes should be:
- evocative, thought provoking, emotive, attention grabbing, add to the story, explain the story
describe the process of contextualising quotes in the writeup
- Quotes can’t stand by themselves in isolation
- Need to be illustrative of research findings
- When reporting quotes:
- Say what the finding is
- Introduce the quote by saying who said it and in what context
- Provide the quote
- Comment on the quote by comparing it to other quotes from the same person, other quotes from different people on the same topic, etc.
