Qualitative Data Analysis Notes

CABS Research Assistance Training Package - Module 6: Qualitative Data Analysis

Lecture Outline

  • Extracting interesting points from raw qualitative data

  • KJ method of qualitative data analysis

Qualitative Data

  • Definition: Qualitative data seeks to develop models and theories to describe and explain the motivations, attitudes, and behavior of people.

  • Key Components:

    • Individual Construal

    • Situational Context

    • Behavior

    • Observation

  • Methodological Approaches:

    • Interview

    • User Journey Mapping

    • Focus Groups

    • Card Sorting

    • Naturalistic Observation

    • Task (or Work) Analysis

    • Contextual Inquiry

Qualitative Data Analysis Caveat

  • Methods taught in this module are aimed at practitioners and are efficient for practical purposes.

  • Caution: These methods are not suitable for academic research or publication, as academic methods are more complex and yield mostly similar outcomes.

  • For those interested in academic qualitative data analysis, consult the book "Qualitative Data Analysis – A Methods Sourcebook".

Extracting Interesting Points from Raw Qualitative Data

Types of Raw Qualitative Data Collected:

  • Audio recordings from interviews

  • Notes from observations

  • Data from diary studies

  • Notes and audio from focus groups and contextual inquiry

  • Photos from observations and diary studies

Process of Interesting Points Extraction:

  1. Extract Interesting Points: Convert raw data into interesting points.

  2. Analyze Interesting Points: Use KJ analysis to form themes from interesting points.

  3. Transition: From raw data to interesting themes.

Categories of Raw Qualitative Data:

  • Raw Analog Data:

    • Audio recordings from interviews, focus groups, and contextual inquiries

    • Video recordings from observations and contextual inquiries

    • Photos from observations and diary studies

  • Raw Textual Data:

    • Notes from interviews, observations, focus groups, and contextual inquiries

    • Open-ended responses from diary studies

Cleaning Raw Qualitative Data

  • Raw qualitative data (e.g., from interviews, focus groups, and observation studies) must be “cleaned” and formatted into interesting points before analysis can begin.

Steps for Extracting Interesting Points

  1. Convert Raw Analog Data into Notes or Transcripts:

    • Essential for audio, video, and photographic data.

  2. Extract Interesting Points from Notes or Transcript:

    • Break down raw qualitative text into smaller, manageable information chunks, termed as 'interesting points'.

  3. Inter-Coder Comparisons of Interesting Points:

    • Ensures data quality and reduces bias.

Detailed Procedures for Data Extraction

Step 1: Convert Raw Analog Data into Notes or Transcripts

  • Raw analog data represents qualitative information physically (e.g., sounds, images).

  • Recommendation: Use transcription software for audio recordings for a fairly accurate initial text from recordings, while editing afterwards for quality assurance.

  • Best Practices:

    • Find a quiet place, use quality headphones.

    • Transcribe at 2x speed if possible.

    • Aim to transcribe verbatim; include stutters and filler words if time allows.

    • Time-stamp responses for reference.

Step 2: Extract Interesting Points from Notes or Transcripts

  • Interesting Points Defined:

    • Behaviors, feelings, thoughts, and opinions expressed by participants.

    • Unexpected quotes and themes that are frequently repeated are noteworthy.

    • Follow the rule of extracting one interesting point per minute on average.

    • Interesting points should be summarized in short phrases (4-6 words).

  • Coding Templates:

    • Templates can trace coded points back to raw data.

  • Recommendation: Utilize two coders to minimize bias in extracting points to enhance data quality.

Example of Interesting Points Extraction

  • Supporting Evidence: Include verbatim quotations or descriptions from audio/video recordings that led to extracting the interesting points.

Inter-Coder Comparisons of Interesting Points

  • Core Purpose: Minimize bias of individual coders and ensure the reliability of the data extracted.

  • Procedure:

    • Both coders compare notes; approximately 80-90% alignment is expected.

    • Standardize wording on interesting points across coders.

    • For differing points (10-20% variability), discuss until agreement is reached, involving a neutral third party if necessary.

KJ Method of Qualitative Data Analysis

  • Developed By: Jiro Kawakita, recognized for synthesizing large quantities of data into manageable thematic chunks through participant collaboration.

  • Procedure Overview:

    1. Find A Space: Can be physical or digital for planning and organizing data.

    2. Assemble Team: Involve participants in data collection and analysis.

    3. Create Cards: Represent interesting points on physical or digital cards.

    4. Sort Cards: Organize based on thematic similarities.

    5. Label Groups: Identify each group of cards.

    6. Regroup: Reorganize based on commonalities.

    7. Team Walk-through of Diagram: Finalize the structure and connections of data.

Practice Activities

  • Activity 1a: Correct the transcript of an interview by comparing it with raw audio to identify errors.

  • Activity 1b: Extract interesting points from corrected transcripts using illustrated templates.

  • Further group work to carry out KJ analysis on Facebook posts related to the qualitative research topic.

Group Dynamics in KJ Method

  • Ensure a collaborative effort from team members to foster creativity and mitigate bias.

  • Finally, conclude with theme definitions while drawing connections between themes to encapsulate the key messages gathered from qualitative analysis.