W3 L1: Qualitative Data Analysis

THEMATIC ANALYSIS

the analytical process

  • involves developing codes, categories and/or themes

  • it begins with transcription and reading - raw data

  • researchers immerse themselves in the data

  • checking, revising, and refining emergent knowledge by returning to data

coding

  • examining emperical material, labelling it and identifying related content across the dataset

  • word or short phrase

multiple approaches

  • grouped theory

  • content analysis

  • conversational analysis

  • conversational analysis

  • narrative analysis

  • thematic analysis

  • etc.

Thematic Analysis

  • approaches that focus on identifying themes

  • minimally organises and describes your data in detail

3 types of TA

  • small q approach

- oriented around coding reliability

  • medium q approach

- based on a structure codebool and qualitative philosophy

  • big q approach

- reflexive approach based on ‘organic’ coding

refelxive thematic analysis

  • Braun & Clarke’s approach to TA

  • first proposed in 2006

  • 2019 renames to Reflexive TA - highlights the emphasis on researcher reflecivity

  • MUST READ BEYOND 2006 PAPER

Flexibility

  • not tied to particular theoretical framework but not atheoretical

  • used to address most questions and most types of qualitative data

  • researcher woll need to make choises

  • be clear about decisions when writing

analytical decisions

  • orientation to data - inducive (bottom-up) or deductive (top-down)?

  • focus on meaning - semantic or latent (more implicit, requires more interpretation from reader)?

  • qualitative framework - experiemental or critical?

  • episstemology & ontology - critical realism or relativisim, constructionism?

6 phases

  1. familiarisation with data

  • decisions about level of detail wanted in analysis

  • reading and notes (thoughts) about data

  • not findings - just reflexion

  • pay equal attention to data

  1. coding data

  • start finsings

  • capture meaning of data

  • represent meaning of what ppts are saying

  • doesn’t need to be Nvivo

  • decide between inductive and deductive code - or combination of both

  • decising between semantic and latent code

  1. generating initial themes

  • what is presented in report

  • final output

  • cluster together codes that have a similar meaning - creating a theme

  • name the themes - organise story in data, need to be very specific

  • the story that the data is telling

  • 3-5 themes (no more)

  1. reviewing and developing themes

  • ask questions:

is this a theme? what is the quality of this theme? what are the boundaries? is there enough data to support this theme? is the data too diverse and wide ranging?

  1. refining, defining, and naming themes

  • name or lable

  • ‘headline’ that captures the central organising concept - gives reader instant insight into meaning of theme

  • avoid topic summaries

  • write short description for each theme - usually at the beginning of the results section

  • draw on quotations

  • themes are actively generated/ developed by researcher - not passively emerge from data

  • subthemes can be used if needed

  1. writting up

  • set of themes are presented

  • select vivid and copelling data

  • relate analysis to research question

  • draw out analytical conclusions across themes

doing reflexive TA

  • messy

  • not linear

  • organic

  • back and forth

semantic - using words used by ppt

latent - use code arojund ‘moral/ good’ implying a meaning