Selective Coding in Qualitative Research

Selective coding is the final coding stage in Strauss and Corbin’s grounded theory, a method for building theory from data. After open coding (labelling data) and axial coding (relating categories), researchers choose a core category. This central concept ties all other categories together and tells the data’s main story.

The key difference from earlier coding stages is integration and refinement. You’re no longer breaking the data into labelled segments (like in open coding) or making initial connections between categories (like in axial coding).

Instead, you’re weaving everything together into a unified theoretical explanation.

The core category acts as a unifying thread, connecting the various categories and subcategories into a cohesive theoretical framework.

The goal of selective coding is to move beyond description and develop a grounded theory that explains the phenomenon under study.

It ensures that the analysis is not merely a collection of disparate themes but rather a coherent and integrated whole, enhancing the theoretical contribution of the research.

When do I start selective coding?

Selective coding, the final stage of grounded theory development, is not a rigidly defined process with a fixed starting point.

It’s an iterative and evolving process that emerges naturally from the preceding stages of open and axial coding.

The decision of when to transition depends on three things. These are the richness of your data, your evolving understanding of the phenomenon, and the emergence of a potential core category.

Here are some indicators that suggest you’re ready to embark on selective coding:

• Saturation of Categories: You’ve reached a point where additional data analysis doesn’t yield significant new categories or insights.
• Clear Relationships Between Categories: The connections between your existing categories have become more apparent, revealing potential causal relationships or patterns of influence. You start to see how different categories might fit together into a larger theoretical framework.
• Emergence of a Potential Core Category: A central, unifying category begins to stand out from the others. This category appears frequently, connects meaningfully to other categories, and offers strong explanatory power for the observed patterns in the data.
• Desire for Theoretical Integration: You feel a growing need to move beyond describing the individual categories and start weaving them together into a cohesive and explanatory theory.

Saturation is easy to mistake for repetition. Saunders et al. (2018) separated its meanings, which helps you judge whether your categories are ready to be integrated.

So ask whether your categories are fully developed, not whether the interviews have started to sound repetitive.

Steps in Selective Coding

Selective coding is the final stage in grounded theory methodology. It focuses on identifying a core category and weaving all other categories around it into a coherent theoretical explanation of the phenomenon under study.

Selective coding is iterative and reflexive. You may need to revisit and revise your core category and the relationships between categories multiple times as your understanding evolves.

This ongoing engagement with the data ensures the development of a robust and grounded theory.

Selective Coding Across Grounded Theory Traditions

Glaser and Strauss (1967) introduced grounded theory but left the coding procedure loosely specified. Later writers filled that gap in different ways. As a result, selective coding means different things in different traditions.

• Strauss and Corbin: Selective coding is the final phase, after open and axial coding. The analyst picks a core category, relates all categories to it and writes the storyline (Strauss & Corbin, 1990, 1998).
• Glaser: Glaser rejected axial coding as forcing. Selective coding follows open coding: once a core category emerges, the analyst codes only for it and related concepts to delimit the study (Glaser, 1978, 1992).
• Charmaz: Charmaz (2006) renamed the phases initial and focused coding. Focused coding uses the most significant or frequent initial codes to sift large amounts of data.

Glaser’s alternative to axial coding is theoretical coding. The analyst picks abstract “coding families”, such as the “Six Cs” (causes, contexts, contingencies, consequences, covariances and conditions), that fit the emerging theory. The steps below follow the Strauss and Corbin sequence.

1. Identifying the Core Category:

A core category (also called a central theme) is the unifying concept of selective coding. It is a high-level, abstract idea that integrates the main themes from the data into one theoretical framework.

A strong core category has two qualities:

• Explanatory power: It explains the patterns in the data and offers insight into why things are happening, not just a description.
• Abstraction: It rises above the details of individual data points, which allows broader generalization and theory development.

Strategies for identifying the core category:

Identifying a core category is iterative. Examine the data and the relationships between categories closely, and revise your choice as your understanding evolves.

• Immerse yourself in the data: Begin by thoroughly reviewing all codes and categories developed during open and axial coding. The goal is to develop a deep understanding of the recurring patterns, connections, and potential explanations embedded in the data.
• Prioritize frequent and prevalent categories: Look for categories that appear consistently across multiple data points and participants. This frequency signals the potential significance of a category and its relevance as a potential core category.
• Look for natural connections and explanatory power: The core category should link naturally to other categories and unify them. It should explain why the patterns occur, not just describe them.
• Seek abstraction for theory development: A strong core category is abstract enough to rise above individual data points. This lets you build a more generalizable theory that applies to other contexts.
• Consider the “empirical puzzle”: Pay close attention to puzzling or unexpected findings in the data. These anomalies often hold the most significant insights and can lead to the identification of a core category that challenges existing assumptions.
• Use visual aids to map relationships: Employ visual tools like mind maps or network diagrams to visualize the relationships between categories and help you identify potential core categories. These visual representations enhance understanding and facilitate the identification of connections.
• Engage in constant comparison: Continuously compare the core category to other categories, refining its definition and scope as your understanding of the data evolves. This iterative process helps ensure the core category’s robustness and relevance.

2. Relate Categories to the Core Category:

Rigorously examine the connections between the core category and other categories, using evidence from the data to confirm the validity of these relationships.

Ask yourself how the categories influence each other and how they contribute to the overall phenomenon?

If the relationship between the core category and other categories is unclear or underdeveloped, the researcher further refines those categories to achieve a more precise and meaningful connection.

This might involve:

• Splitting: Dividing a broad category into more specific subcategories.
• Merging: Combining conceptually similar categories.
• Adding Detailed Descriptions: Providing richer explanations of how each category relates to the core category.

Explore any areas where the existing categories don’t fully account for the phenomenon. Develop new categories to capture these missing dimensions.

Strategies for systematically relating categories:

• Validate relationships through evidence: Support the connections between the core category and other categories with concrete examples and evidence from the data. This grounding in the data ensures the credibility and trustworthiness of the theoretical explanation.
• Employ dense description: Provide rich and detailed descriptions of how each category relates to the core category. Clearly articulate the connections, influences, and potential causal relationships between them. These descriptions form the building blocks of the theoretical framework.
• Refine categories for clarity and precision: Adjust existing categories to ensure they connect meaningfully to the core category. Use strategies like splitting, merging, or adding detailed descriptions to achieve conceptual clarity and refine the framework’s structure.
• Fill conceptual gaps with new categories: As you explore the relationships between categories, remain open to identifying new themes or concepts that might emerge. Develop new categories to capture these additional dimensions, enriching the theoretical framework’s explanatory power.
• Consider temporal relationships: Pay attention to the chronological order or sequence of events and how they relate to the core category. These temporal connections contribute to building a process model that explains the unfolding of the phenomenon.

3. Writing a Coherent Theoretical Explanation

Selective coding culminates in the development of a grounded theory that’s clearly articulated and grounded in the data.

Three parts matter: components, requirements and presentation.

The core category serves as the central pillar of your theoretical framework, while other categories act as supporting beams and structural elements.

The researcher integrates these major categories around the core category to create a comprehensive explanation of the phenomenon.

A well-developed grounded theory should:

• State your theory in a few words or sentences, providing a concise representation.
• Define the boundaries of your theory: what it explains and what it doesn’t explain.
• Summarize the core category, its relationships with other categories, and the overall storyline.
• Provide a comprehensive explanation that includes variation rather than assuming a “one-size fits all” answer to the research question.

Grounded theories can be presented in different formats:

• As a basic process with distinct phases.
• As a story or narrative that explains how participants create their own understanding and meaning of reality.

Each element, from core components to presentation choices, contributes to a coherent and well-grounded theory.

Strategies for building a coherent theoretical explanation:

• Organize categories hierarchically: Structure the major categories around the core category in a clear and logical hierarchy. This framework makes the relationships between categories readily apparent and enhances the theory’s coherence.
• Articulate the how and why: Move beyond describing the what of the phenomenon. Explain the processes, mechanisms, and contextual factors behind how and why it occurs.
• Support claims with rich data excerpts: Use vivid and illustrative data excerpts to support the claims and propositions within the theoretical explanation. These excerpts ground the theory in the participants’ experiences, enhancing its persuasiveness and trustworthiness.
• Validate the theory against the data: Continuously check the theoretical explanation against the data, ensuring it accurately reflects the lived experiences and perspectives of the participants. This iterative validation process strengthens the theory’s rigor and grounding in the data.
• Maintain a neutral and objective stance: Present the findings and theoretical explanation in a clear, objective, and unbiased manner. Avoid using emotionally charged language or inserting personal opinions into the analysis. This neutrality reinforces the study’s credibility.

Selective Coding Example: Adjusting to Chronic Illness

Imagine an interview study of how people adjust to a chronic illness diagnosis. This is an illustration, not a published study.

After open and axial coding, the analyst has three categories:

• Controlling information about the illness: concealing the diagnosis, passing as well and managing disclosure.
• Living within new limits: rationing energy and organising life around the body.
• Preserving a valued self: resisting a spoiled identity, performing the former self and identity disruption.

Now the analyst needs one category that explains the rest. An earlier memo, the analyst’s running note on emerging ideas, suggested that people conceal partly to protect identity, not just privacy.

So the analyst tests “preserving a valued self” as the core category. Concealing information and rationing energy can then be read as strategies for protecting that self.

The analyst checks each link against the interview data. Where the data disagree, the analyst revises it. The core category now organises the whole analysis.

Selective Coding FAQs

What is the difference between selective coding and axial coding?

• Axial coding focuses on developing and refining categories and exploring relationships between them.
• Selective coding goes a step further by identifying a core category and integrating all other categories and subcategories around it.

Think of axial coding as building the foundation and selective coding as constructing the entire building around a central pillar.

The stages also differ in what they produce. Axial coding uses Strauss and Corbin’s “paradigm model” to link a category to its causal conditions, context, intervening conditions, action strategies and consequences. Selective coding then integrates all categories around one core category.

The table compares the three stages of the Strauss and Corbin sequence.

Stage	Main task	In practice
Open coding	Breaking the data apart	Label segments of data with codes and group them into categories.
Axial coding	Putting the data back together	Relate categories to their sub-categories using the paradigm model.
Selective coding	Integrating and refining	Choose a core category, relate all other categories to it and write the storyline.

Why is selective coding important?

Selective coding is essential for:

• Developing a grounded theory: It allows researchers to move beyond description and develop an explanatory theory grounded in the data.
• Creating a focused and coherent analysis: It ensures that the analysis is not just a collection of disparate themes but a cohesive and integrated whole.
• Enhancing the theoretical contribution: By identifying a core category and relating all other findings to it, the analysis gains depth and provides a more substantial contribution to knowledge.

Can I use existing theories during selective coding?

This is a debated topic in grounded theory. Glaser generally discourages the use of pre-existing theories, while Strauss and Corbin are more open to it.

Theoretical sensitivity is a key concept in grounded theory. It is the researcher’s ability to recognize and analyze data while remaining open to emerging concepts.

While remaining grounded in your data, you can use existing theories to guide your thinking and help you make sense of your findings.

Do not force your data into pre-existing theoretical frameworks. The goal is a theory grounded in your data, not validation of an existing one.

If you do use existing theories, be transparent about how you used them and how they contributed to your analysis.

---

References

Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative analysis. Sage.

Glaser, B. G. (1978). Theoretical sensitivity: Advances in the methodology of grounded theory. Sociology Press.

Glaser, B. G. (1992). Basics of grounded theory analysis: Emergence vs. forcing. Sociology Press.

Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research. Aldine.

Saunders, B., Sim, J., Kingstone, T., Baker, S., Waterfield, J., Bartlam, B., Burroughs, H., & Jinks, C. (2018). Saturation in qualitative research: Exploring its conceptualization and operationalization. Quality & Quantity, 52(4), 1893–1907. https://doi.org/10.1007/s11135-017-0574-8

Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Sage.

Strauss, A., & Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd ed.). Sage.