03 · What You Need to Know
Evaluating the Credibility of Early Data Saturation Claims
Is There a Minimum Number of Interviews Before Saturation Can Occur?
There is no universally established minimum number of interviews that must be completed before any form of saturation can be considered.
However, this does not mean that every early saturation claim is methodologically defensible.
The amount of data required depends on the research aim, participant characteristics, interview depth, analytical strategy, and form of saturation being assessed.
A narrowly defined descriptive study may identify a recurring set of issues relatively quickly. A study seeking detailed interpretations of diverse experiences or developing conceptual explanations may require substantially more information.
The relevant question is whether the sample supports the intended analytical purpose, not whether it meets or falls below a familiar numerical threshold.
What Does It Mean to Claim Saturation After Five Interviews?
The statement is incomplete unless the researcher specifies what was saturated.
For example, the researcher might mean that no additional descriptive codes appeared during the fifth interview.
Alternatively, the researcher might claim that all relevant meanings had been sufficiently explored or that conceptual categories had been fully developed.
These are substantially different claims.
A small dataset may support a preliminary judgment about recurring descriptive content without providing adequate evidence for a comprehensive interpretation of the phenomenon.
Researchers should therefore define the analytical target before assessing whether saturation has occurred.
Why Is the Difference Between Code and Meaning Saturation Important?
Hennink, Kaiser, and Marconi (2017) distinguish code saturation from meaning saturation.
Code saturation concerns identifying the range of relevant issues represented by the data. Meaning saturation concerns developing a richer understanding of those issues, including their dimensions, contextual conditions, and variation.
| Analytical question |
Code saturation |
Meaning saturation |
| What is being assessed? |
Whether additional interviews identify new relevant codes. |
Whether additional interviews substantially deepen understanding of existing issues. |
| What might early repetition indicate? |
Recurring descriptive topics. |
Not necessarily sufficient interpretive depth. |
| What may still be missing? |
Rare or context-specific issues. |
Nuance, contradictions, processes, and conceptual relationships. |
| What should the researcher examine? |
Evidence that additional interviews contribute few new relevant codes. |
Evidence that existing issues are understood sufficiently for the analytical aim. |
For example, five teachers may all identify academic integrity as a concern when using generative AI. That repetition does not necessarily explain how teachers interpret integrity, what decisions they make, or why their practices differ.
The meaning of data saturation for sample-size planning must therefore be specified rather than assumed.
What Does Empirical Research Say About Early Saturation?
Guest, Bunce, and Johnson (2006) investigated thematic saturation in a particular qualitative interview dataset and reported that major themes became apparent relatively early in their analysis.
Hennink, Kaiser, and Marconi (2017) subsequently demonstrated that identifying codes and understanding their meanings can require different amounts of data.
Hennink and Kaiser (2022) reviewed empirical studies examining saturation and found that sample requirements varied across research conditions.
These studies provide useful methodological evidence, but they should not be interpreted as establishing that five, six, or twelve interviews are universally sufficient.
Their findings are conditional on the study populations, analytical approaches, interview procedures, and operational definitions of saturation.
Why Might Saturation Appear to Occur Prematurely?
An apparent absence of new information may result from the data collection process rather than genuine analytical adequacy.
For example, participants may provide repetitive responses because the interview questions are narrow, leading, or insufficiently probing.
Recruitment may also concentrate on participants with nearly identical experiences, leaving important contextual variation unexplored.
Premature analytical closure can occur when researchers begin with a limited set of expected categories and overlook material that challenges those expectations.
These possibilities should be investigated before interpreting repetition as saturation.
Can Poor Interview Questions Create False Saturation?
Yes. Questions that invite brief agreement or disagreement can produce similar answers without generating sufficient information.
Suppose a researcher asks five teachers, "Do you believe generative AI creates academic integrity concerns?" All five answer yes.
The repetition may appear to support a recurring code, but the responses reveal little about the nature of those concerns or their influence on practice.
More informative questions might ask participants to describe particular incidents, explain their decisions, and reflect on the consequences.
Improving interview quality can reveal variation that was not apparent in the initial accounts.
The relationship between interview quality and qualitative sample size is therefore central to evaluating early saturation claims.
Can a Highly Homogeneous Sample Produce Early Repetition?
Yes. Participants with similar experiences may discuss similar issues, especially when they work within the same institution or professional context.
Such repetition can be informative when the research aim deliberately concerns that specific group.
However, it becomes problematic when the researcher draws conclusions about broader populations or contexts that were not adequately represented.
For example, five teachers from one academic department may provide a focused account of that department's assessment practices but insufficient information for conclusions about disciplinary differences across a university.
The specificity of the sample should be evaluated against the scope of the claims.
Does a Highly Specific Research Aim Make Early Saturation More Plausible?
Potentially. A narrowly bounded research question may require information about a relatively limited set of experiences.
For example, a study describing how members of a small, specialized committee implemented one particular procedural change may involve a limited eligible population and a focused descriptive objective.
Even in that setting, researchers should distinguish the completeness of participant coverage from the adequacy of the analysis.
Interviewing nearly everyone involved may provide valuable evidence, but it does not automatically demonstrate that every relevant meaning or process has been understood.
The scope of the research aim remains important when evaluating whether early repetition is meaningful.
How Can Researchers Examine Whether Important Variation Is Missing?
Researchers should compare the achieved sample with the characteristics and experiences required by the research question.
For example, if the study concerns teachers' responses to institutional AI policies, relevant variation may include differences in disciplinary practice, assessment responsibilities, and direct experience implementing the policies.
A sample consisting entirely of enthusiastic AI adopters may omit teachers who reject or cautiously restrict AI use.
Additional purposive recruitment may be needed to investigate those missing perspectives.
However, researchers should not attempt to represent every imaginable subgroup. Variation should be selected because it matters to the intended analysis.
What Role Do Negative or Discrepant Cases Play?
Negative cases challenge an emerging interpretation and can reveal conditions under which an apparent pattern does not hold.
Suppose the first five participants describe AI as reducing assessment-development workload. A later participant reports that verifying AI-generated content increased their workload substantially.
The contrasting account may reveal a distinction involving assessment type, disciplinary expectations, or institutional support.
Its analytical importance cannot be dismissed merely because it is uncommon.
Researchers should examine whether early saturation claims reflect adequate engagement with possible contradictions rather than only repetition of dominant experiences.
Can Saturation Be Assessed Using Interview Batches?
In compatible coding-oriented approaches, researchers can examine how successive interview batches contribute new relevant information.
Guest, Namey, and Chen (2020) proposed a structured method for assessing thematic saturation using a base set of interviews, subsequent runs, and an operational threshold for new information.
Such procedures can make saturation decisions more transparent.
However, a numerical threshold should be justified in relation to the study and should not be treated as a universal rule.
With only a few interviews, the evidence supporting a stable pattern of diminishing new information may be particularly limited.
What Should a Researcher Document Before Claiming Early Saturation?
A credible assessment should explain the analytical procedure and the evidence supporting the claim.
| Assessment area |
Evidence to examine |
Possible concern |
| Saturation definition |
Clear specification of codes, meanings, or conceptual categories. |
The term is used without an operational meaning. |
| Interview quality |
Detailed accounts, examples, and appropriate probing. |
Repetition results from superficial questioning. |
| Participant variation |
Coverage of experiences relevant to the research aim. |
Important perspectives are absent. |
| Analytical progression |
Memos or records showing how additional interviews contributed. |
No evidence of concurrent analysis. |
| Contrasting accounts |
Consideration of discrepant experiences and alternative interpretations. |
Only confirming evidence is emphasized. |
| Stopping rationale |
Explanation of why further data were unlikely to add consequential information. |
Recruitment ended solely because of time or access constraints. |
This is a practical review framework, not a validated saturation scoring instrument.
Its purpose is to make the evidence behind the decision visible and open to methodological scrutiny.
What If Saturation Was Claimed Only After Data Collection Ended?
Retrospective assessment is not automatically invalid, but researchers should distinguish it from a stopping decision informed by concurrent analysis.
If recruitment ended after five interviews because only five participants were available, the researcher should not imply that an ongoing saturation assessment caused recruitment to stop.
The existing material can still be evaluated for its adequacy to the research question.
However, the report should accurately describe the sequence of recruitment and analysis and acknowledge any limitations.
Retrospective claims should not conceal practical constraints or imply a procedure that was never conducted.
Is Early Saturation Appropriate for Grounded Theory?
Grounded theory traditions commonly emphasize iterative analysis and theoretical sampling.
Theoretical saturation concerns whether conceptual categories and their relationships have been sufficiently developed.
Identifying recurring topics after a few interviews does not establish that a grounded theory has achieved theoretical saturation.
Researchers should examine the specific grounded theory tradition, the development of conceptual categories, and whether further theoretical sampling is needed.
An early claim requires particularly careful explanation when the study seeks to develop a substantive theoretical account rather than describe a limited set of experiences.
What If the Study Uses Reflexive Thematic Analysis?
In reflexive thematic analysis, saturation may be inconsistent with the approach's assumptions about theme development.
Braun and Clarke (2021) question the use of saturation as a sample-size rationale for thematic analysis, particularly when themes are understood as actively constructed through interpretation.
Researchers using this approach should not attempt to defend five interviews merely by asserting that no new themes emerged.
Instead, they should justify the sample through research scope, participant relevance, dataset richness, and the depth of the resulting analysis.
The question of whether saturation fits the qualitative methodology must be resolved before judging the credibility of an early saturation claim.
Can Information Power Provide an Alternative Assessment?
Information power offers a different framework for considering qualitative interview sample adequacy.
Malterud, Siersma, and Guassora (2016) emphasize study aim, sample specificity, established theory, dialogue quality, and analysis strategy.
A small sample may provide substantial information when the aim is focused, participants are highly relevant, and interviews support intensive analysis.
However, information power does not automatically validate a sample of five participants.
The researcher must explain how the actual material supports the intended analytical claims.
Information power should not be used as a convenient replacement label for an unsupported saturation claim.
Should Researchers Always Recruit More Participants After an Early Saturation Claim?
Not automatically. Additional recruitment should address a clearly identified informational or analytical need.
If the existing interviews are superficial, improving interview technique or conducting follow-up interviews may be more useful than simply increasing the number of participants.
If important perspectives are absent, targeted recruitment may be necessary.
If the study is intentionally idiographic or case-focused, deeper analysis of the existing material may be more appropriate than expanding the sample.
The decision should follow the research aim and methodology rather than an assumption that more participants are always better.
What If Additional Recruitment Is Impossible?
When recruitment cannot be expanded, researchers should evaluate what the available data can reasonably support.
They may need to narrow the research question, reduce unsupported subgroup comparisons, or present the findings as a limited exploratory account.
However, changes to the research aim should be scientifically justified and documented rather than made solely to conceal recruitment difficulties.
The researcher should acknowledge missing perspectives, limited analytical depth, or other consequential limitations.
A modest but transparent contribution is preferable to an expansive claim supported by insufficient evidence.
How Should Reviewers Evaluate a Claim of Saturation After Five Interviews?
Reviewers should avoid rejecting the study solely because the sample is small. They should also avoid accepting saturation merely because the authors use the term.
The review should examine whether saturation is appropriate to the methodology, whether its definition is clear, and whether the analytical evidence supports the claimed stopping point.
Particular attention should be given to the research aim, sampling boundaries, interview quality, and the breadth of the conclusions.
If the justification is inadequate, reviewers can request a more precise account of the sampling and analytical procedures.
Where additional data collection is not feasible, the appropriate response may be narrower claims and clearer limitations rather than an unsupported assertion of saturation.
How Should Early Saturation Be Reported?
A research article should state the form of saturation assessed, how interviews were analyzed, and what evidence supported the decision.
For example, a coding-oriented study might describe how successive interviews were reviewed for new relevant codes and whether existing categories required further elaboration.
The report should also explain the sampling strategy and the variation represented in the achieved sample.
Researchers should distinguish an analytical stopping decision from recruitment ending because of resource constraints.
Statements such as "saturation was reached after five interviews because responses were repetitive" provide insufficient methodological detail on their own.
