Module 4: Grounded Theory and Narrative Analysis Study Notes

Overview of Grounded Theory

• Origins: Grounded Theory is an increasingly popular qualitative approach that emerged in the mid to late $1960s$. It is considered a relatively new approach to understanding research data collection.
• Creators: The methodology was created and coined by two individuals: Barney Strauss and Anson Glazer.
• Primary Purpose: The central goal is theory development and the generation of new knowledge. It is specifically used to:
  • Discover problems existing within a social scene.
  • Understand how individuals handle those problems.
  • Formulate and test ideas.
  • Redevelop propositions until a definitive theory is established.
• Theory Basis: Grounded theory is unique because it seeks to define not just the problem, but the specific ways humans deal with social issues via social interactions.

Philosophical Foundations of Grounded Theory

• Underlying Epistemologies: According to Glazer and Strauss, there are two key approaches to knowledge that inform this theory:
  • Pragmatism: This concept posits that knowledge production is a matter of perspective. Warrants or claims to truth depend on consensus among members of a community of inquiry.
  • Symbolic Interactionism: This is the concept that social interactions are meaningful and that shared meanings amongst groups of people are influential in society.
• Core Focus: The emphasis is on a shared understanding of the world. Data is collected from groups of people to identify these shared meanings.

The Grounded Theory Research Process: 9 Steps

• Step 1: Determine Initial Research Question(s): Identify a phenomenon to explain. These guide initial recruitment and analysis but evolve as the researcher learns more from the data.
• Step 2: Recruit and Collect Data Using Theoretical Sampling: This is an iterative process. Unlike quantitative research, recruitment criteria are not fully predetermined.
• Step 3: Break Transcripts into Excerpts Using Open Coding: Verbatim transcripts are broken into individual pieces called excerpts (sentences or small paragraphs related to one theme).
• Step 4: Group Excerpts into Codes Using Open Coding: Excerpts representing the same central concept are grouped under a "code" (a label or tag).
• Step 5: Group Codes into Categories Using Axial Coding: Compare codes to find connections and group them into broader categories.
• Step 6: Analyze More Excerpts Using the Constant Comparative Method: A cyclical process of comparing new data against existing codes/categories to find contradictions, expansions, or support.
• Step 7: Continue Data Collection and Analysis Until Theoretical Saturation: The point where additional data no longer adds or expands upon existing codes.
• Step 8: Define the Core Category Using Selective Coding: Connect all categories under one central thesis or core category that forms the basis of the theory.
• Step 9: Construct the Grounded Theory: Use the core category, coded data, and memos to write a short, clear statement of the theory, define its boundaries, and provide evidence.

Data Collection and Sampling in Grounded Theory

• Purposive Sampling: A deliberate selection of individuals relevant to the research question.
  • Researchers must define sample characteristics that provide relevant data (e.g., if studying mothers, only collect data from mothers).
  • Researchers must explicitly state criteria for selection to increase credibility and quality.
• Theoretical Sampling: A process where previous analysis influences subsequent data collection decisions.
  • Start with a small group, analyze, and let the findings guide who to recruit next.
  • Considered an essential characteristic of a complete grounded theory study.
• Methods of Data Collection:
  • Interviews: The most common method.
  • Focus Groups.
  • Participant Observation.
  • Review of Records: Diaries, documents, images, videos, and photos.

Grounded Theory Analysis and Coding Phases

• Nature of Analysis: It is an iterative and recursive process where each stage informs the next. This differs from quantitative structures that are set from the start.
• The Constant Comparative Method: The core of grounded theory analysis. Researchers compare excerpts from the same person, different people, different experiences, or different time points.
• Transcription: Data must be transcribed (verbatim for interview data) into a written format for analysis.
• Memos: "Notes to self" where researchers record their train of thought, reflections, interpretations, and contradictions found in data. These are the building blocks for the final theory.
• Coding Levels:
  • Open Coding: Breaking data into excerpts and assigning initial labels (tags).
  • Axial Coding: Creating links between open codes to form categories (e.g., grouping "trouble sleeping" and "panic attacks" under "anxiety").
  • Selective Coding: Determining the single core category/statement of meaning that encapsulates all other findings (e.g., "Access to lockdown-suitable housing mitigated Covid-19 anxiety").
• Outcomes of Constant Comparison:
  • Contradiction: New data contradicts a code; requires adjusting the code or conducting more theoretical sampling.
  • Expansion: New data adds detail/facets; indicates more collection is needed.
  • Support: New data adds no new information; results in Theoretical Saturation.

Evaluation of Grounded Theory: Advantages and Disadvantages

• Advantages:
  • Real-world Representation: Derives theories directly from participants in natural settings rather than labs.
  • Data Connection: Findings are tightly linked to the participants' actual experiences.
  • New Discoveries: Strong inductive method for novel theories without preconceived hypotheses.
  • Structured Analysis: Offers specific strategies (open, axial, selective) to keep the process analytical.
  • Reduced Confirmation Bias: The interwoven nature of collection and analysis protects against proving preconceived beliefs.
• Disadvantages:
  • Difficulty in Recruitment: Iterative recruitment requirements and evolving criteria make finding participants challenging.
  • Time-Consuming: The researcher cannot know in advance how much data or how many rounds of sampling are required.
  • Complexity of Analysis: Data is analyzed on a rolling basis; keeping track of constant comparisons can be overwhelming.
  • Cost: May require specialized data analysis software to manage the high volume of information.

Narrative Analysis: Definitions and Data Sources

• Definition: An approach used to understand how participants construct stories and narratives from personal experience. "Narrative" is synonymous with "story."
• Dual-Layered Interpretation:
  1. The participant interprets their life through a narrative.
  2. The researcher interprets how the participant constructed that narrative.
• Data Sources: Journals, letters, conversations, autobiographies, and in-depth interview transcripts.
• Transcription Detail: Emphasizes verbatim transcription including pauses, filler words (e.g., "um"), and straight utterances, as these form the narrative structure.

Types of Narratives in Research

• Topical Stories: Restricted stories about a specific moment in time (e.g., an answer to a single question about childhood). Includes plot, characters, and setting but not a full life history.
• Personal Narratives: Extended accounts of life from long interviews or a series of interviews over time (e.g., collecting stories over the course of a week).
• Entire Life Story: A complete biography constructed from a collection of interviews, observations, documents, and historical artifacts.

The Narrative Analysis and Transcription Process

• Coding Approaches:
  • Deductive Coding: Approaching analysis with a preset structure of codes.
  • Inductive Coding: No preset structure; led entirely by the data. (Narrative analysis often uses a blend of both).
• Narrative Blocks: Unlike other methods that break data into excerpts, narrative analysis keeps entire blocks of text intact for analysis.
• Analysis Steps (The Dell Framework):
  • Step 1: Inductively code narrative blocks based on similar life events (e.g., "deciding to have children").
  • Step 2: Read all blocks with the same code to note similarities and differences.
  • Step 3: Create a nested story structure using deductive codes.
  • Step 4: Collate life events by story structure (e.g., focusing only on the "Orientation" of different stories).
  • Step 5: Compare across different story structures to see how meaning construction differs.
  • Step 6: Tell the core narrative, observing how stories relate or diverge.

Wendy Pattison’s Six Story Structure Codes

• Abstract: The core thesis or summary of the story.
• Orientation: The time, place, situation, and characters involved.
• Complicating Action: The sequence of events or the plot.
• Evaluation: How the storyteller comments on the meaning of the life event.
• Resolution: The outcome of the story.
• Coda: The formal ending of the story.

Evaluation of Narrative Analysis

• Advantages:
  • Depth: Reveals in-depth detail of specific life events.
  • Historical Significance: Reflects the historical context of the time the narrative was captured.
  • Lived Experience of Minorities: Useful for small sample sizes (even $1\text{ to }3$ individuals), allowing marginalized voices to be heard.
• Disadvantages:
  • Researcher Embedding: Requires the researcher to have a deep internal understanding of the topic to represent the experience accurately.
  • Time-Consuming: Both interviewing and analyzing detailed data take significant time.
  • High Subjectivity: Focuses on lived experience rather than broader social movements.
  • Memory Reliability: Relies on subject memory; may require triangulation (cross-referencing with other documents) for accuracy.

Qualitative Interviewing Formats and Techniques

• Unstructured Interview:
  • Format: Opening statement, then spontaneously generated questions based on the flow.
  • Strength: Maximum flexibility to pursue information as it emerges.
  • Weakness: Less systematic; difficult and time-consuming to analyze.
• Structured Interview:
  • Format: Approximately $20\text{ questions}$ asked in the exact same word order and sequence for every person.
  • Strength: Reduces variation between interviewers; good for comparing many subjects/time points.
  • Weakness: Cannot pursue interesting topics that arise in the moment.
• Semi-Structured Interview:
  • Format: An interview guide with $4\text{ to }5\text{ key questions}$ to be covered flexibly. Serves as a checklist rather than a rigid script.
  • Strength: More systematic than unstructured while remaining conversational and deep.
  • Weakness: Difference in wording/sequencing can reduce comparability of responses.

Transcribing and Analyzing Interview Data

• Verbatim Transcription Conventions:
  • Reproduce everything, including vocal expressions like "um," "err," and "dunno."
  • Pauses: $\text{...}$ denotes a pause of $3\text{ seconds}$ or more.
  • Stressed Words: Typed in bold.
  • Shouted Words: CAPITALISED.
  • Whispered Words: Italicized.
• Analysis Strategies:
  1. Identifying Themes: Underlining sections of the transcript in pencil based on the meaning of words.
  2. Coding Themes: Writing short code names on the right-hand margin. Use a separate "Codes" sheet to define themes and note line numbers.
  3. Memoing: Using a separate "Memos" sheet to note ideas about relationships between themes.
  4. Creating 'Trees': Organizing themes into a hierarchy. The "trunk" is the most important theme, while "branches" represent subsidiary or expanded themes.