Thematic Analysis: From Study Design to Write-Up (L4Rema)

Introduction to Thematic Analysis

  • Definition: Thematic analysis (TA) is described as an accessible, flexible, and increasingly popular method of qualitative data analysis.

  • Purpose: It is a method for systematically identifying, organizing, and offering insight into patterns of meaning (themes) across a data set (Braun and Clarke 2012).

  • Braun and Clarke (2006/2012) Approach:

    • Developed a standardized 6-phase framework for researchers.

    • Emphasized the flexibility of the approach, allowing it to be adapted to various research contexts.

    • Provides a useful structure for studies involving multiple researchers.

    • Foundational principle: "The analysis produces the answer to a question."

Constructing the Research Question

  • Research Problem: There must be a identified GAP in existing research and a corresponding NEED to conduct the new study.

  • Research Question Definition: The question should aim specifically to address the research gap to help fulfill the identified need.

  • Quantitative Framework (PICO): A method primarily for formulating strong quantitative research questions.

    • P: Population

    • I: Intervention

    • C: Control

    • O: Outcomes

    • Example Question: "(P) In school-age children (I) what is the effect of a school-based physical activity program (O) on a reduction in the incidence of childhood obesity (C) compared with no intervention?"

  • Qualitative Framework (PEO): Often differs from quantitative approaches by focusing on exploration.

    • P: Patient, Population, or Problem (The disease, condition, and demographic).

    • E: Exposure (Exposure to a condition, risk factor, screening, rehabilitation, or service).

    • O: Outcome or Themes (Experiences, attitudes, feelings, quality of life).

    • Example Question: "Experiences of using steam inhalation in patients with chronic sinusitis."

      • P=Patients with Chronic sinusitisP = \text{Patients with Chronic sinusitis}

      • E=Using steam inhalationE = \text{Using steam inhalation}

      • O=ExperiencesO = \text{Experiences}

  • Mixed Methods Framework (SPIDER): Designed for qualitative and mixed-methods research.

    • S: Sample (The group of people being studied).

    • PI: Phenomenon of Interest (Topic of the research, such as an intervention).

    • D: Design (Techniques used for data gathering: focus groups, interviews, observations).

    • E: Evaluation (The outcome of the study).

    • R: Research Type (Qualitative methodology: Phenomenology, Ethnography, Grounded Theory, Case Study).

    • Example Question: "What are the barriers experienced by clinicians that lead to the reluctance to recommend CBT in practice? A qualitative interview study."

      • S=CliniciansS = \text{Clinicians}

      • Pi=Recommending CBTPi = \text{Recommending CBT}

      • D=InterviewsD = \text{Interviews}

      • E=Barriers to recommending CBT in practiceE = \text{Barriers to recommending CBT in practice}

      • R=Phenomenology (experiences)R = \text{Phenomenology (experiences)}

Sampling and Qualitative Methodology

  • Sampling Techniques:

    • Purposive Sampling: Selecting participants based on specific required characteristics (e.g., women, over age 40, registered at a specific GP practice, with no history of CVD).

    • Convenience Sampling: Selecting participants based on accessibility and availability (e.g., students current present in a lecture).

  • Data Collection Methods: Interviews, focus groups, etc.

    • Interview Guide: Quality data requires a strong guide consisting of open-ended questions.

    • Structure of Interview Questions:

      • Question 1: Should be open and easy to answer to get the participant talking (e.g., "Can you tell me how you first realised you may be diagnosed with…?").

      • All questions should include prompts to encourage deeper responses.

      • Final Question: Always ask if there is anything else the participant wishes to add.

The 6 Phases of Thematic Analysis (Braun and Clarke)

  • Inductive vs. Deductive: TA is typically inductive (themes are generated directly from the data). Deductive analysis involves testing an existing theory, which is not the standard application for TA.

Phase 1: Familiarizing Yourself With the Data
  • Goal: Immerse yourself in the data.

  • Activities: Reading and re-reading transcripts, listening to audio recordings, watching videos of interviews/focus groups, and taking initial notes.

Phase 2: Generating Initial Codes
  • Definition: The systematic analysis of data through coding.

  • Every "meaningful" section of text is assigned a code.

  • Crucial Rule: Not every piece of text needs to be coded. If it does not relate to the research question, it is ignored.

  • Phrases used for codes can be refined (e.g., changing "Happy" to "Happiness at diagnosis").

  • Example: For the text "New colleagues… might be needing to look at it more," an initial code might be "New staff."

Phase 3: Searching for Themes
  • Goal: Shift from individual codes to broader themes.

  • A theme captures something important about the data relative to the research question.

  • Codes are grouped into meaningful clusters; occasionally, one code may become an overarching theme.

  • Example Structure:

    • Theme: Usefulness of prompts.

    • Subthemes: Useful for inexperienced practitioners; Support for decision; A reminder tool.

Phase 4: Reviewing Potential Themes
  • Goal: Themes are reviewed against the coded data and the entire data set.

  • Researchers ask:

    • Is there another way codes could be grouped?

    • Do themes adequately answer the research question?

    • Are the quotes representative of each code/theme?

  • Example: Mapping "Friendly staff" under the theme "Satisfaction with clinic."

Phase 5: Defining and Naming Themes
  • Goal: Clearly state what is unique and specific about each theme.

  • Criteria for Good Themes:

    • Singular focus (they do not try to do too much).

    • Related but non-overlapping (no repetition).

    • Directly address the research question.

  • Example: A subtheme titled "Assistance in persuading patients" describes how GPs use prompts to help explain to patients why antibiotics may not be necessary.

Phase 6: Producing the Report
  • Goal: Create a journal article or dissertation that Tells a "compelling story."

  • Order matters: Themes should connect logically. Researchers should rearrange themes to ensure a meaningful flow.

  • Example Revision: Changing the order from "Use of therapy -> Experiences of trauma -> Receiving diagnosis" to "Experiences of trauma -> Receiving diagnosis -> Use of therapy."

Writing up a Qualitative Research Paper

  • Introduction:

    • Identify the research problem and the specific need/gap.

    • State the research question clearly.

  • Methods:

    • Participants: Provide exhaustive detail.

    • Procedure: Explain the sampling technique and why it was chosen.

    • Materials: Describe tools like the interview guide and its development.

    • Analysis: Detail the stages used (e.g., citing Braun and Clarke).

  • Results:

    • Present an overview of themes and subthemes.

    • Provide descriptions for each theme and support them with verbatim quotes.

    • Note: Do not discuss implications or refer to previous evidence in this section.

  • Discussion:

    • Summarize key findings and relate them to the research question.

    • Highlight the most important and prevalent themes.

    • Compare findings to previous literature, discuss implications, note limitations, and reach a conclusion.

Worked Example: McDermott et al. (2010)

  • Study Title: "Developing a computer delivered, theory based intervention for guideline implementation in general practice."

  • Research Problem:

    • Need: GPs are not adhering to guidelines for Respiratory Tract Infections (RTI) and stroke prevention.

    • Gap: Electronic prompts can help, but they have varying success and the use of behavior change theory in their design is a missing element.

  • Methods:

    • 33 face-to-face interviews with GPs in South England and London.

    • Prompts were presented to GPs during semi-structured interviews.

  • Overarching Question: "What factors may be involved in GPs' decision to use electronic prompts for RTI and stroke prevention guidelines during a consultation?"

  • Key Themes Identified:

    • Perceptions of role of prompts: GPs rejected "enforcement" but accepted "choice/control."

    • Patient outcomes: Assistance in persuading patients.

    • Prescriber differences: Usefulness for inexperienced staff.

    • Accessibility: Usability and information presentation.

    • Acceptability: Credibility of sources.

  • Conclusion: Acceptability of computer-delivered prompts is higher when practitioners view them as valuable tools that support rather than mandate their practice.