Introduction to Qualitative Data Analysis (PSY5014M)
Foundations of Qualitative Data Analysis (QDA)
Definition and Purpose: Qualitative Data Analysis is defined as a process of deconstructing and reconstructing evidence through purposeful interrogation and critical thinking. It entails exploring similarities, differences, and meanings to generate categories and themes, with the ultimate goal of producing a meaningful interpretation of the data (Bazeley, 2021; Lyons & Prior, 2021).
Nature of the Process: Qualitative analysis is characterized as intense, engaging, challenging, non-linear, contextualized, and highly diverse. It is not a process that occurs in a vacuum; rather, it takes place within the specific context of a research project and requires the researcher to adopt a particular ‘qualitative stance’ or worldview (Wertz et al., 2011; Lyons & Coyle, 2021).
Post-Collection Transition: Research moves from data collection to analysis. Data can take many forms, including news articles, social media posts (e.g., Twitter/X), academic assessments, and visual materials.
Case Study Example: Children's Geographies (Lehman-Frisch et al., 2012):
Study Title: "'Draw me your neighbourhood': a gentrified Paris neighbourhood through its children’s eyes."
Methodology: Researchers used drawings and interviews from 4th-grade children in a socially mixed, gentrified neighborhood of Paris (Batignolles).
Objective: To reflect on the methodological contribution of drawings in analyzing children’s representations of their city and studying how gender and social background impact their experiences.
Specific Examples from Findings:
Marc (Figure 1): Drew a map-like representation including the Parc Cardinet Batignolles and Rue Cardinet.
Lilian (Figure 2): Focused on specific buildings and a "SHOPING" area.
Sophie (Figure 3): Represented a dense neighborhood.
Olivia (Figure 4): Captured landscape and architecture.
Olivier (Figure 5): Depicted a shopping district including a "tabac" and "boulangerie."
Salih (Figure 6): Illustrated a lively area with bars and shops.
Shelihane (Figure 7): Provided a very rough, symbolic drawing of her building courtyard. The analysis noted her struggle to express herself and categorized her household as "too small" (a studio where she lived with her mother, a cleaner). Her neighborhood was reduced to the exterior faade of her home.
Conceptual Frameworks: "Big Q" vs. "Small q"
Researchers generally categorize qualitative data analysis into two distinct philosophical orientations:
Big Q (Fully Qualitative Research):
Characteristics: Processes are flexible, interpretive, and subjective/reflexive.
Philosophical Stance: Rejects the idea of objectivity and context-independent truths.
Meaning: Emphasizes the contextual and situated nature of meaning (Braun & Clarke, 2021; Coyle, 2021).
Small q (Qualitative Positivism):
Characteristics: Uses qualitative techniques (like coding) but remains within a hypothetical-deductive framework.
Focus: Primarily concerned with the accuracy and reliability of coding.
Methodology: Utilizes structured codebooks or coding frames applied to data, often with multiple coders working independently to ensure consistency.
Overview of Qualitative Analytical Approaches
There are numerous approaches to QDA, each with different epistemological assumptions and procedural requirements:
Thematic Analysis (TA):
Definition: An umbrella term for identifying patterns of meaning (themes) across a dataset (Braun & Clarke, 2021).
Epistemology: Researchers select a position that fits their specific aims (e.g., description of experience vs. deconstruction of meaning).
Sampling: Typically purposive sampling of at least six individuals (though variable).
Outcome: A set of illustrated and interpreted themes that answer the research questions.
Discourse Analysis (DA):
Definition: An umbrella term for analyzing the use and functions of talk and text within social interaction (Wiggins, 2017).
Epistemology: Constructionist; views language as constructing social and psychological worlds rather than reflecting pre-existing realities.
Sampling: Preference for naturally occurring data; requires enough text to discern variety in discursive forms.
Outcome: An analysis of interpretive repertoires and functions, focusing on what texts are "doing."
Interpretative Phenomenological Analysis (IPA):
Definition: An idiographic approach focused on detailed examinations of personal lived experience and how participants make sense of their world (Smith & Eatough, 2021).
Epistemology: Phenomenological with a realist commitment; language is seen as reflecting experience.
Sampling: Usually small, homogenous samples (up to 10 people) with first-hand experience of a phenomenon.
Outcome: A set of themes attentive to both the group and the individual perspective.
Other Notable Approaches: Grounded Theory, Content Analysis, Discursive Psychology, Narrative Analysis, and Conversation Analysis.
Analytical Pluralism
Definition: The application of more than one qualitative analytical method to a single dataset.
Benefits:
Provides a more comprehensive answer to research questions.
Avoids privileging a single approach.
Enables rich, complex understandings and multiple possibilities for interpretation.
Variations: Can involve using multiple methods, data sources, theories, or researchers (Clarke et al., 2015; Lyons & Prior, 2021).
Common Practices and Norms in QDA
Regardless of the specific method, several foundational norms apply to qualitative analysis (Wertz et al., 2011):
Meaning-Making: Analysis is inherently a process of creating meaning.
Data Immersion: Starts with open (re)reading of data with sensitivity to context.
Intentionality: Focus on participant intentions and meanings.
Refinement: Interpreting requires iterative cycles of checking, revising, and refining by returning to the raw data.
Evidence-Based Claims: Knowledge claims must be supported with evidence (e.g., quotes or excerpts).
Transparency: Providing a transparent accounting of all procedures used.
The Role of Reflexivity and Positionality
Definition of Reflexivity: Critical reflection on the research process and the researcher's own role. It acknowledges that the researcher's position affects the research and challenges the idea of objective knowledge production (Berger, 2015; Burr, 2015).
Types of Reflexivity:
Personal Reflexivity: How the researcher's background, identity, interests, and views shape the research.
Discursive Reflexivity: Detailed attention to the researcher's role within the specific research process behaviors.
Epistemological Reflexivity: How theoretical assumptions, research questions, and data collection methods shape the findings (Cooper & Burnett, 2006).
Researcher Positioning: Factors include gender, ethnicity, age, sexual orientation, immigration status, linguistic tradition, beliefs, biases, and emotional responses to participants' experiences.
Impact of Positionality: Influences access to the field (who joins the study), the researcher-researched relationship (what is shared), and the final findings/conclusions.
Strategies for Demonstrating Reflexivity:
Explicitly stating beliefs and experiences.
Using first-person language (, ).
Maintaining a research journal or diary for "self-supervision."
Using data triangulation or peer review.
Representing the "Other"
Definitions: Organizations of power often construct specific groups as "Other," including women, people with mental health conditions, LGBTQ+ individuals, economically disadvantaged groups, those with HIV/AIDS, Black and minority communities, and people with disabilities.
Insider vs. Outsider Status:
Insider: May facilitate access but risks identifying too closely with participants, complicating analysis.
Outsider: May lack contextual knowledge; however, apparent "sameness" between an insider and participant can be illusory due to diverse internal group experiences.
Strategies to Mitigate "Othering":
Participant validation: Asking participants to evaluate the researcher's account.
Listening to how participants describe the "research group."
Creating a dialogue between researcher and participant to ensure no single account is privileged (Coyle, 2021).
Establishing Trustworthiness and Rigor
Adler (2022) notes that quantitative research often has a "facade" of intrinsic trustworthiness due to its systematic appearance, whereas qualitative research must actively work to establish credibility.
Dimensions of Trustworthiness (Lincoln & Guba, 1985):
Credibility (Quantitative Analogue: Internal Validity): Confidence in the truth value. Strategies: Member checking, researcher triangulation.
Transferability (Quantitative Analogue: External Validity): Applicability to other settings. Strategies: Thick description of data, contextual information about participants.
Dependability (Quantitative Analogue: Reliability): Consistency and coherence. Strategies: Transparent reporting, audio-recording, and transcription.
Confirmability (Quantitative Analogue: Objectivity): Neutrality; findings reflect participants rather than researcher bias. Strategies: Inclusion of raw data, reflexivity, and triangulation.
The Shift to "Rigor": There are calls to replace the term "trustworthiness" with "rigor," although researchers rejecting scientific epistemologies may find empirical evaluation tools ill-fitting (Adler, 2022).
Computer-Assisted Qualitative Data Analysis Software (CAQDAS)
Function: Tools for managing qualitative data, particularly during coding and annotating large datasets.
Available Software:
Nvivo: Available at York St John University (YSJ).
Atlas.ti
Transana
LIWC
ELAN: Open access.
Caveat: Software helps with data management but does not "do the analysis" for the researcher. The interpretive work remains the responsibility of the human researcher.
AI and QDA: emerging issues regarding how Artificial Intelligence might automate or influence these processes.
Take Home Messages
Approach selection depends on the research question and purpose.
Reflexivity is an ongoing critical appraisal of the researcher's role.
Qualitative rigor is evaluated through dimensions like credibility and transferability, which differ from quantitative metrics.
CAQDAS tools manage data but do not replace human interpretation.
Questions & Discussion
Knowledge Check: What are the key differences between 'Big Q' and 'Small Q' approaches to qualitative data analysis?
Response Summary: Big Q is fully qualitative, subjective, and flexible, rejecting objective truths; Small Q is qualitative positivism, using qualitative methods within a deductive framework focused on coding accuracy.