Theoretical Saturation In Grounded Theory

Theoretical saturation in grounded theory refers to the point in the research process when gathering additional data about a theoretical category doesn’t reveal any new properties or provide further insights into the emerging grounded theory.

At this point, “theoretical completeness” is considered to have been reached.

Grounded theory research involves simultaneous data collection and analysis. Researchers collect data, analyze it through coding, and use the insights gained to guide further data collection. This iterative process continues until theoretical saturation is reached.

The idea dates to Glaser and Strauss (1967) and The Discovery of Grounded Theory. There, theoretical sampling lets emerging concepts decide who or what to study next. Sampling continues until further data reveal no new properties of a category.

The indicators of saturation are not always clear-cut. Judging them can be subjective.

Researchers need to exercise their analytical skills, theoretical sensitivity, and judgment to determine when saturation has been reached.

Discussing emerging findings with colleagues and seeking feedback can help researchers assess the robustness and completeness of their analysis. So can engaging in reflexivity, which means staying aware of how their own assumptions and biases shape the codes they generate.

Conceptual Density in Grounded Theory

Conceptual density refers to the level of abstract understanding and interconnectedness among categories and concepts within a grounded theory.

Corbin and Strauss treat dense category development as one condition for saturation.

A theory with high conceptual density is typically characterized by four features:

1. Abstract Understandings: A conceptually dense theory goes beyond simply labeling or categorizing data. It generates abstract concepts that capture the underlying patterns and processes at play. For example, instead of simply noting that participants reported “feeling stressed,” a conceptually dense theory might identify a more abstract concept like “managing uncertainty” to explain the participants’ experiences and actions.
2. Interconnected Categories: A conceptually dense theory doesn’t just present a list of concepts. It explains how these concepts relate to each other, forming a coherent and integrated framework. This often involves identifying a core category, the one to which all other categories relate. It integrates them and highlights the central process or phenomenon under investigation.
3. Theoretical Coding: Integrating relevant theoretical codes from existing bodies of knowledge can strengthen the explanatory power of a grounded theory and enhance its conceptual density. For instance, a grounded theory about coping mechanisms in chronic illness might draw on existing theories of stress and resilience to provide a deeper understanding of the observed patterns.
4. Explanation, Not Just Description: Grounded theory aims to generate theories that explain social processes, not just describe them. This distinction is crucial because a conceptually dense theory needs to move beyond surface-level observations to offer insights into the “why” and “how” of the phenomenon.

Achieving Conceptual Density

Grounded theorists can employ several strategies to enhance the conceptual density of their research:

1. Cultivate Abstract Thinking:

• Move Beyond Descriptive Coding: During initial coding, you’ll likely generate descriptive codes that closely reflect the data. As you progress, challenge yourself to develop more abstract concepts that capture the underlying meanings and processes. For instance, instead of just coding instances of “patients expressing frustration,” consider developing a more abstract category like “navigating system barriers” to encapsulate the underlying experience.
• Look for Connections and Patterns: As you code, constantly compare data, codes, and categories to identify relationships, patterns, and potential causal connections. Ask yourself: How do these codes relate? What seems to be driving these actions? What are the consequences of these interactions?

2. Embrace Memo-Writing:

Memos can be short notes about codes and categories or longer reflections on emerging themes and relationships in the data.

They provide a valuable audit trail, help guide theoretical sampling, and contribute to the development of the grounded theory.

• Record Theoretical Insights: Use memos to capture your evolving understanding of the data. Document your thoughts about how codes and categories relate to each other, possible explanations for observed patterns, and potential connections to existing theories.
• Explore Contradictions and Inconsistencies: Memos are also a place to grapple with contradictions or inconsistencies you encounter in the data. Exploring these tensions can lead to richer and more nuanced theoretical insights.
• Use Memos to Guide Analysis: Review your memos regularly to identify recurring themes, gaps in your understanding, and areas for further theoretical sampling.

3. Engage in Theoretical Sampling:

• Refine Your Focus: Theoretical sampling involves selectively choosing participants or data sources that will provide the most insightful information to develop your emerging theory. As your analysis progresses, you’ll identify areas where you need more data to clarify concepts, explore relationships, or test emerging hypotheses.
• Don’t Be Afraid to Adjust Your Sample: Your initial sampling strategy might need to be adjusted as your theory evolves. Be prepared to revise your ethics applications as needed to accommodate new participant groups that emerge as theoretically important.
• Seek Diversity and Variation: Theoretical sampling often involves seeking out participants or data sources that offer diverse perspectives or represent different variations of the phenomenon you are studying. This helps ensure that your theory is comprehensive and captures the complexity of the social world.
• Example: A researcher exploring patient experiences with chronic pain might begin with open sampling, interviewing patients with diverse conditions and pain management approaches. As analysis reveals concepts like pain intensity, coping mechanisms, and social support, the researcher might then theoretically sample patients experiencing specific types of pain (e.g., acute vs. chronic) or utilizing different coping strategies. This iterative process continues until categories reach theoretical saturation.

4. Constant Comparison of Data:

Dense category development relies on the constant comparative method, where researchers continuously compare new data with existing categories and their properties.

Constant comparison also signals when to stop. Researchers check each new piece of data against existing categories and concepts, asking whether it adds any new insight or property.

The method is systematic and iterative. Researchers compare data, codes, and categories to identify patterns, relationships, and theoretical insights.

Through constant comparison, researchers can refine their theoretical understanding, challenge existing assumptions, and ensure that the emerging theory is grounded in the data.

By continuously comparing data, researchers can assess whether the existing sample is sufficient to reach saturation.

Sample size follows the data. If new data keep adding to the theory, the sample may need to grow. A larger sample captures the full range of perspectives and experiences related to the phenomenon.

If new data reveal no new properties or insights for the emerging categories, the sample is sufficient. Data collection can stop.

5. Employ Theoretical Coding:

• Draw on Existing Knowledge: Theoretical codes are concepts, perspectives, or theoretical frameworks borrowed from existing bodies of knowledge. They are used to analyze and interpret your data, enriching the explanatory power of your grounded theory. For example, if you’re developing a grounded theory about decision-making in healthcare, you might draw on existing theories of risk perception or shared decision-making to provide a theoretical lens for interpreting your findings.
• Select Codes Strategically: Only use theoretical codes that “earn their way” into the analysis by demonstrating a clear fit with the data. Don’t force your data into pre-existing theoretical frameworks.
• Use Theoretical Coding to Explore Relationships: Theoretical codes can be particularly helpful for understanding how the categories you’ve developed relate to each other and for articulating the underlying processes at play.

6. Avoid Premature Closure:

• Resist the Urge to Rush to Conclusions: Grounded theory is an iterative process. Allow your theory to emerge gradually through cycles of data collection, coding, memo-writing, and theoretical sampling.
• Stay Open to New Insights: Be willing to revise your concepts, categories, and even your core category as you gather new data and refine your understanding.
• Embrace Theoretical Sufficiency: The goal is to reach a point of theoretical sufficiency, where your categories adequately account for new data without needing constant modifications. Recognize that this is a judgment call based on your interpretation of the data.

Grounded theory is a dynamic process, so treat these six strategies as iterative and interconnected.

Theoretical Sufficiency

Dey (1999) suggests “theoretical sufficiency” as a more fitting substitute for “saturation.”

On this view, data collection does not wait for absolutely no new information. That standard is impractical. It ends when the researcher understands the phenomenon well enough to build a theory.

This viewpoint underscores that complete knowledge is unattainable. The objective is adequate understanding for theory development.

Theoretical sufficiency implies that the researcher has gathered enough data to understand the phenomenon under investigation comprehensively.

It does not require every single detail or variation. It requires a level of knowledge that supports a well-grounded and insightful theory.

The shift from data saturation to theoretical sufficiency aligns with reflexive thematic analysis. Braun and Clarke (2021) argue that saturation concepts do not fit this approach, because meaning is generated through interpretation rather than excavated from the data.

• Meaning is not simply “discovered” within data but is actively constructed through the researcher’s ongoing engagement and interpretation.
• The process of analysis is iterative and reflexive, with codes and themes constantly being refined and reinterpreted in light of new data and evolving understanding.

Given this dynamic process, a fixed endpoint of “no new information” becomes problematic. New insights depend on the researcher’s evolving analytical lens.

Therefore, theoretical sufficiency favors a flexible approach to data collection that values depth and richness of understanding over accumulating data points.

Corbin and Strauss’ Three Elements of Theoretical Saturation

Corbin and Strauss, in their work on grounded theory, define theoretical saturation as the point in the research process when three specific elements have been fulfilled:

1. No New Data: Newly collected data add no new insights or properties to a category.
2. Dense Categories: Categories have rich, well-defined properties and dimensions.
3. Validated Relationships: The links between categories are well established and supported by the data.

No New or Relevant Data Emerge

As data collection and analysis progress, newly collected data may stop providing new insights or properties for existing categories. Interviews, observations, and other data sources yield information that the developed categories and their properties already capture.

Essentially, the data become repetitive and redundant. That indicates the category has been fully explored and elaborated.

Data triangulation supports this judgment: multiple data sources, such as interviews, field notes, and documents, converge on similar findings.

This does not imply an absolute end to new information. It suggests that further exploration of this area is unlikely to be fruitful.

Dense Category Development

In Corbin and Strauss’s formulation, category development is dense insofar as all of the paradigm elements are accounted for. Categories are dense when researchers have developed them thoroughly, with rich descriptions of their properties and dimensions.

Properties are the characteristics or attributes of a category. Dimensions represent the variation or range within a property.

A dense category has well-defined properties and dimensions that capture the complexity and variability of the phenomenon. Researchers should provide thick descriptions of the categories. Ample evidence from the data should support them.

This element suggests that the analysis has explored the dimensions and nuances of each category. The result is a rich understanding of its properties and its place within the broader theoretical framework.

Well-Established Relationships

Theoretical saturation is not just about individual categories. It also depends on understanding how categories relate to and influence one another within the overall theory.

These relationships form the core of the emerging theory. Saturation is reached when they are stable, well supported by data, and provide a coherent and meaningful understanding of the phenomenon.

Validating the relationships ensures that the emerging theory is internally consistent and offers a comprehensive explanation of the phenomenon under investigation.

Corbin and Strauss define saturation specifically for grounded theory.

On their account, reaching saturation means researchers can feel confident that the theory captures the essence of the studied experience. It also gives them a useful framework for understanding it.

Saturation of Core Category

Corbin and Strauss emphasize the importance of identifying a core category that integrates the other categories and represents the central phenomenon under investigation.

It acts as a unifying thread. It frames the relationships between the concepts and processes identified in the data.

Saturation of this core category, along with its related subcategories, is vital for a comprehensive theory.

The core category often sits at the top of a hierarchy of categories. Subcategories branch out below it. They explain different aspects of the phenomenon in more detail.

Corbin and Strauss add a caveat. Saturating the core category alone is not enough. Its subcategories must also be saturated to give a comprehensive and detailed understanding of the phenomenon.

Ultimately, the core category and its saturated subcategories are integrated into a coherent theoretical framework that explains the “why” and “how” of the phenomenon under investigation.

Critical Evaluation of Theoretical Saturation

Theoretical saturation is grounded theory’s standard stopping rule. Researchers still disagree about what it means and how to show it has been reached.

Why Theoretical Saturation Is Contested

Researchers often conflate theoretical saturation with data saturation, the point at which new data merely repeat what is already present. The two differ.

Theoretical saturation concerns the development of categories, not the tallying of repeated content. Fixed interview counts therefore sit awkwardly with it.

Guest et al. (2006) tracked data saturation across sixty in-depth interviews with women in two West African countries. Saturation occurred within the first twelve interviews, although basic elements for metathemes appeared as early as six.

Low (2019) challenges the usual definition itself. Equating saturation with “no new information” gives researchers no guidance on how to recognize that point.

It is also a logical fallacy, Low argues, because new theoretical insights keep appearing as long as data collection continues. Low offers a more pragmatic definition in its place.

Contemporary Research

Recent work has tightened the vocabulary of saturation and tested how many interviews it takes in practice.

Saunders et al. (2018): Four Models of Saturation

For grounded theory, coding stops when categories, with their properties and dimensions, are fully developed through theoretical sampling. Repetition alone is not enough.

Simulating Theoretical Saturation

Van Rijnsoever (2017) used simulations to estimate the sample size needed to reach theoretical saturation. He defined saturation as observing every code in a population at least once.

He compared three sampling scenarios. Under random chance, sources are drawn by probability.

Under minimal information, each new source adds at least one new code. Under maximum information, each adds as many as possible.

Saturation depended more on how likely codes were to be observed than on how many codes existed. The two deliberate scenarios were significantly more efficient than random chance, but yielded fewer repetitions of each code to validate the findings.

He recommends following the minimal-information scenario.

Code Saturation Versus Meaning Saturation

Hennink et al. (2017) analyzed 25 in-depth interviews to compare two ways of judging saturation. Code saturation, when the range of issues has been identified, came at nine interviews.

Meaning saturation, a richly textured understanding of those issues, took 16 to 24. The gap is wide.

Put simply, code saturation means researchers have “heard it all,” but meaning saturation means they “understand it all,” according to the authors.

Hennink and Kaiser (2022) later reviewed 23 articles that tested saturation empirically or by statistical modeling. Studies using empirical data reached saturation within 9 to 17 interviews or 4 to 8 focus groups, particularly with homogeneous populations and narrowly defined objectives.

Multi-country research, metathemes, and code-meaning saturation needed larger samples.

Most of these studies assessed code saturation in thematic analysis. Their numbers guide sampling, but they do not show that grounded-theory categories are complete.

References

Aldiabat, K. M., & Le Navenec, C. L. (2018). Data saturation: The mysterious step in grounded theory method. The Qualitative Report, 23(1), 245-261. https://doi.org/10.46743/2160-3715/2018.2994

Braun, V., & Clarke, V. (2021). To saturate or not to saturate? Questioning data saturation as a useful concept for thematic analysis and sample-size rationales. Qualitative Research in Sport, Exercise and Health, 13(2), 201-216. https://doi.org/10.1080/2159676X.2019.1704846

Dey, I. (1999). Grounding grounded theory: Guidelines for qualitative inquiry. Academic Press.

Francis, J. J., Johnston, M., Robertson, C., Glidewell, L., Entwistle, V., Eccles, M. P., & Grimshaw, J. M. (2010). What is an adequate sample size? Operationalising data saturation for theory-based interview studies. Psychology & Health, 25(10), 1229-1245. https://doi.org/10.1080/08870440903194015

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

Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18(1), 59-82. https://doi.org/10.1177/1525822X05279903

Hennink, M., & Kaiser, B. N. (2022). Sample sizes for saturation in qualitative research: A systematic review of empirical tests. Social Science & Medicine, 292, Article 114523. https://doi.org/10.1016/j.socscimed.2021.114523

Hennink, M. M., Kaiser, B. N., & Marconi, V. C. (2017). Code saturation versus meaning saturation: How many interviews are enough? Qualitative Health Research, 27(4), 591-608. https://doi.org/10.1177/1049732316665344

Low, J. (2019). A pragmatic definition of the concept of theoretical saturation. Sociological Focus, 52(2), 131-139. https://doi.org/10.1080/00380237.2018.1544514

Rowlands, T., Waddell, N., & McKenna, B. (2016). Are we there yet? A technique to determine theoretical saturation. Journal of Computer Information Systems, 56(1), 40-47. https://doi.org/10.1080/08874417.2015.11645799

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. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd ed.). Sage.

van Rijnsoever, F. J. (2017). (I can’t get no) saturation: A simulation and guidelines for sample sizes in qualitative research. PLOS ONE, 12(7), Article e0181689. https://doi.org/10.1371/journal.pone.0181689