Qualitative Data Analysis Notes

Recording Interviews

  • Documenting interviews is crucial; audio recording is common, with video recording increasingly used.
  • Recordings allow revisiting the interview, creating transcripts, and enabling later analysis.
  • The primary purpose of recording is to allow full attention to the interviewee during the interview.
  • Taking notes can break eye contact, causing missed cues.
  • The goal is to create a naturalistic interview setting.

Recording Tips

  • Use high-quality, reliable equipment, potentially accessing university resources or grants.
  • A separate microphone placed closer to the interviewee is beneficial.
  • Be cautious of cables to avoid tripping hazards and potential loss of recording.
  • Pre-test equipment before and at the interview location.
  • Maintain a comfortable distance from the interviewee, respecting proxemics.
  • Ensure the interviewee's voice is audible by adjusting volume.
  • Avoid bright lights that may cause discomfort.
  • Control the environment as much as possible to minimize noise and distractions.

Soundscapes

  • Background sounds provide rich contextual information.
  • Soundscapes can be a valuable source of data.
  • Three main types of soundscapes:
    • Human-generated sounds:
      • Electromechanical (traffic, trains, planes).
      • Physiological (coughing, sneezing).
      • Controlled (recorded music).
      • Incidental sounds.
    • Non-human, living sounds:
      • Animals (dogs barking).
      • Insects.
      • Birds.
    • Non-biological sounds:
      • Wind, rain, thunder.

Transcriptions

  • A transcript is a verbatim written record of the interview.
  • The transcription can be done yourself or outsourced.
  • Transcription technology is increasingly available, including free online packages.
  • Types of transcripts:
    • Gisted transcript (summary).
    • Basic level transcription (verbatim).
    • Advanced transcription (including ums, ahs, pauses, intonation, and volume using symbols).
  • Elements to include in a transcript:
    • What people said.
    • Non-verbal behavior (in brackets).
    • Other sounds.
    • Comments on group interaction.
  • Benefits of transcribing:
    • Common practice.
    • Relatively easy, especially with technology.
    • Reasonable accuracy.
    • Enhanced data analysis through familiarity.
    • Ability to give full attention during the interview.
  • Potential errors:
    • Mishearing words (e.g., "not" changing meaning).
    • Confusing similar-sounding words (e.g., "adopt" and "adapt").
    • Missing commas or semicolons.
    • Jargon or unfamiliar vocabulary.

Using Transcripts

  • Import data into software for organization and analysis.
  • Remove identifiers to protect privacy, keeping a separate file linking pseudonyms to real names.
  • Clean the data by removing unhelpful elements, but avoid over-cleaning to retain potentially relevant information.
  • Produce summary tables to organize data, similar to literature review tables.

Qualitative Data Analysis

  • Identify, explore, and give meaning to patterns in data to address study questions.
  • An ongoing, adaptable, and emergent process.
  • Four main types of qualitative analysis:
    • Iterative.
    • Enumerative.
    • Grounded theory.
    • Very subjective types.

Iterative Data Analysis

  • A repetitive process, like a dance with the data.
  • Involves noticing, gathering information, sorting, reflecting, and informing future observations.
  • Analytic induction:
    • Start with a question.
    • Generate a hypothesis.
    • Examine data.
    • If data supports hypothesis (no deviant cases), confirm; otherwise, revise the question or hypothesis.

The General Inductive Approach

  • Also known as conventional thematic analysis.
  • Identifies core meanings from interactions or interviews to address study questions.
  • Seven-step process:
    1. Reading:
      • Read the transcript slowly and carefully multiple times.
    2. Coding:
      • Label meaningful text segments, sentences, or phrases.
      • Coding is an iterative process.
      • Fracture the data into parts.
    3. Categorizing Codes:
      • Grouping codes into categories, which become themes.
      • Themes are abstract units that combine codes.
    4. Describing the Theme:
      * Summarize the themes
    5. Connect Themes:
      * Search for relationships between the themes. Connect to the study question.
    6. Interrogate:
      • Investigate how each of your themes vary across different groups based on age, gender, ethnicity, what have you.
    7. Interpret:
      • Interpret the themes in the context of the literature.
  • In vivo coding: using participants' own words as codes.

Coding Considerations

  • Independent vs. Collaborative Coding:
    • Independent coding allows identifying individual perspectives.
    • Collaborative coding, involving multiple researchers comparing codes, useful if you are a post positivist or a pragmatist.
  • Manual vs. Computer-Assisted Coding:
    • Computers aid data management but do not provide answers.
    • Personal preference dictates the method.

Example of Codes/Themes

  • Raw date: "My heart starts to beat hard and irregularly, my limbs are shaking and my mouth gets dry, I think I'm going to die".
  • Extracted codes: irregular heartbeat, shaking limbs, dry mouth.
Identifying Codes and Themes
  • Strategies for identifying codes and themes:
    • Use participants' own words (in vivo coding).
    • Identify metaphors.
    • Note repetitions.
    • Examine relationships and connectors (e.g., "and," "but").
    • Consider transitions between ideas.
    • Be aware of existing literature.
    • Look for what's present and absent (marked and unmarked text).

Cutting and Pasting

  • Physically moving the texts and sorting them

Worked Example

  • Patient story: "The doctor smiled at me and with no false pride for her selfless concern for my well-being she thanked me for my gift”.
  • Line-by-line coding: doctor smiled (friendliness), no false pride (humility), selfless concern (altruism), thanked me for my gift (gratitude).
  • Grouping codes: these codes can be grouped under the theme of virtue (stable traits of good character).

Describing themes

  • Link back to the study questions.

Levels of Interpretation

  • Basic Interpretation: Simply restating what was said.
  • Gentle Interpretation: Adding a little interpretation.
  • Deeper Interpretation: Looking for hidden meanings that may be more risky because of over interpretation, but also more interesting as well.

Over and Under Interpreting

  • Risk of seeing patterns that don't exist (apophenia).
  • Risk of failing to pick up meanings that are actually there.

Types of Analysis

  • Directed by pre existing understanding in literature theory: deductive, top-down.
  • Thematic Analysis: interpreting and looking at the meaning and significance of the content and the context.
  • Content Analysis: an analysis of what was said mainly and describing, describing manifest.
  • Conventional Thematic Analysis: bottom up or inductive (meaning emerge out of the data.)

Saliency versus Frequency

  • Not to equate Theme with recurring codes because a code that is not recurring can still be important. Salient (conspicuous, important, still thematic, based on importance.)

Conventional vs Reflexive Thematic Analysis

  • Conventional: realist, pragmatist, post-positivist.
  • Reflexive: constructionist.
    • Little attention to reflexivity. It asks you to reflect on what you have learned and provide reasons

Enumerative Approach

  • Semi, quasi statistical
  • Counting the number of key words or phrases.

Grounded Theory

  • Start with an open mind, let theory come out of the data. Rooted in the data

Framework Approach

  • Produces a framework that meets information needs of clients and support healthcare practice. It's inductive and deductive (Predetermined or Emergent Codes)

Phenomenology

  • Pure phenomenology: essential nature
  • Interpretative phenomenology: Dialogue, for example, during interviews
  • Investigative approach: language in use

Reporting and Evaluating Qualitative Research

  • Common criticisms of qualitative research:
    • Lacks reproducibility.
    • Lacks generalizability (but has transferability).
    • Subjective.

Transferability

  • Qualiitative studies do not generate generalizible findings via statisical inference.

  • Instead, the researcher provides enough information to the reader for them to decide, for the reader to decide whether they believe that your findings can be transferred to their situation
    Report in a timely manner

  • Can report written text/tables/graphs, can also have visual/photographs art poetry.

    • Who did the research
    • Study Am
    • Methodology
    • What/when/where/who

Reporting Guidelines

  • COREC:Consolidated Criteria for Reporting Qualitative Research (a 32 item checklist for reporting interview analysis)
    • Standards that you are Expected to follow.

Critical Appraisal

  • What is the quality of the work.
  • Provide a balanced evaluation of the strengths and Limitations of what was done.
  • CASP (Critical Appraisal Skills Programme): check list
  • Principles: Qualities, Qualitative, should be careful through and as precise as we can be
  • Trustworthiness: credibility, (transferability, dependability, confirmability. (these are individual things that you can do to assess the quality of the work).
  • Respect
  • Thick Description
  • Prolonged Engagement
  • Audit trail (keep records of what they have/document and processes along the pathway)
  • Triangulation
  • Fairness
  • Demonstrated Reflexivity
  • Member Checking
    • Debriefing
  • Skeptical Peer Review (feedback)
  • Does the other person have similar codes
  • Congruence meaning everything fits
  • Relevance (Health Significance)
  • Novelty (is this new)
  • Connectedness
  • Coherence (fitness and usefulness is really import!
  • Connoisseurship and aesthetic appreciation: How does rather than do a checklist tick tick tick tick, what's your overall impression) It means look at appearances, the aesthetics.