03 · What You Need to Know
Distinguishing Statistical Calculation From Qualitative Sample-Size Justification
Why Do Quantitative Studies Often Calculate Sample Size?
Many quantitative studies seek to estimate population parameters or test statistical hypotheses.
For example, a researcher may want to estimate the percentage of teachers using generative AI within a specified margin of error or detect a difference in achievement scores with a specified statistical power.
Sample-size formulas can be derived from assumptions about sampling distributions, outcome variability, effect sizes, confidence levels, and statistical tests.
These formulas address particular statistical objectives. They do not establish the number of participants needed to understand experiences, develop interpretive themes, or investigate social processes through qualitative analysis.
Applying them to a qualitative interview study without a corresponding statistical estimand creates a mismatch between the calculation and the research objective.
Why Does Qualitative Research Usually Use a Different Justification?
Qualitative studies often seek to understand how participants interpret experiences, how practices operate within particular settings, or how social meanings and processes develop.
The adequacy of the sample depends on the information available for these analytical purposes.
A participant with direct experience of the phenomenon may contribute detailed, contextually relevant material. Another participant may provide a different perspective that challenges an emerging interpretation.
Their contributions cannot be reduced to interchangeable units of statistical information without changing the assumptions underlying the inquiry.
Consequently, qualitative sampling commonly emphasizes relevance, depth, contextual variation, and analytical adequacy rather than a specified probability of rejecting a null hypothesis.
What Is the Difference Between a Sample-Size Calculation and a Justification?
Statistical sample-size calculation
A numerical procedure derived from specified statistical assumptions and objectives, such as estimating a population proportion or testing an effect.
Qualitative sample-size justification
A methodological explanation of why the planned or achieved participants and data are appropriate for the research aim and analytical approach.
Both require reasoning and transparency. Neither is inherently more rigorous merely because one produces a numerical result.
A statistical formula is valuable when its assumptions match the research objective. A qualitative justification is valuable when it explains how the sample supports the intended interpretation.
Should You Use Slovin's Formula for Qualitative Research?
Generally, no. The commonly circulated expression known as Slovin's formula is often presented as a way to calculate sample size from population size and a chosen margin of error.
However, its use does not establish qualitative sample adequacy, and its assumptions are frequently left unspecified in applied research.
For example, knowing that a university employs 800 teachers does not determine how many teachers must be interviewed to understand their experiences of assessment design.
The relevant questions concern which teachers have experienced the phenomenon, what perspectives are needed, and how the material will be analyzed.
Researchers should not apply a population-based formula merely because the institution requires a numerical sample target.
Does Qualitative Research Require Statistical Power Analysis?
Not ordinarily when the study is purely qualitative and does not test a statistical hypothesis.
Statistical power concerns the probability of rejecting a specified null hypothesis under a particular alternative and statistical model.
Qualitative thematic interpretation, phenomenological analysis, and many forms of grounded theory do not ordinarily involve such a hypothesis test.
Therefore, conventional power analysis does not directly determine the number of interviews needed for those purposes.
However, a mixed-methods study may include a quantitative component requiring its own power or precision calculation. The qualitative component should then be justified separately according to its purpose.
Does the Absence of a Formula Mean Sample Size Can Be Arbitrary?
No. Rejecting an inappropriate formula does not remove the obligation to justify recruitment.
A researcher who selects twelve participants solely because twelve is convenient has not necessarily established methodological adequacy.
The sample should reflect the study's aim, the characteristics of participants, the intended analytical strategy, and the information required to support its claims.
Researchers should explain why the proposed sample is reasonable and how they will determine whether the achieved data are adequate.
The broader process of determining an appropriate qualitative sample size requires these judgments even when no statistical formula is used.
Can Information Power Replace a Sample-Size Formula?
Information power provides a framework for reasoning about sample adequacy in qualitative interview studies.
Malterud, Siersma, and Guassora (2016) argue that sample requirements depend on the information participants provide in relation to the research aim.
Their framework considers study aim, sample specificity, established theory, dialogue quality, and analysis strategy.
These dimensions help researchers assess whether a smaller or larger sample is likely to be appropriate.
However, information power is not a mathematical substitute that produces an exact participant count.
Researchers should use it to construct a defensible methodological argument, not to assign arbitrary numerical scores to qualitative characteristics.
The application of information power to qualitative sample-size planning is most useful when its assumptions are compatible with the study's analytical approach.
Can Data Saturation Be Used Instead of a Calculation?
In some qualitative approaches, saturation provides a criterion for evaluating whether additional data collection is likely to contribute substantially new information.
However, saturation is not a single universally defined concept.
Researchers may distinguish code saturation, meaning saturation, and theoretical saturation, each involving different analytical expectations.
Hennink, Kaiser, and Marconi (2017) distinguish the identification of codes from the fuller understanding of their meanings.
Consequently, a claim that no new codes appeared does not necessarily demonstrate that the analysis has achieved sufficient interpretive depth.
Moreover, saturation is not appropriate for every qualitative methodology.
Researchers should specify how saturation will be understood and assessed rather than invoke it as a generic replacement for statistical calculation.
Does Saturation Provide a Number Before Data Collection?
Not by itself. Saturation is commonly evaluated through the relationship between accumulating data and emerging analysis.
Researchers can use previous methodological evidence to inform an initial recruitment range, but the actual adequacy assessment depends on the study.
Guest, Bunce, and Johnson (2006) investigated saturation within a particular interview dataset. Hennink and Kaiser (2022) reviewed empirical evidence concerning saturation sample sizes.
These studies provide useful methodological information but do not establish a universally sufficient number of interviews.
Their findings should be interpreted in relation to the research aim, sample characteristics, data collection method, and definition of saturation.
Is Saturation Appropriate for Reflexive Thematic Analysis?
Its appropriateness is contested, particularly within reflexive thematic analysis as developed by Braun and Clarke.
In this approach, themes are understood as interpretive constructions developed through the researcher's engagement with the data.
The assumption that themes exist as a finite set that can be completely discovered may conflict with this orientation.
Researchers using reflexive thematic analysis should therefore avoid automatically claiming saturation merely because it is expected in a thesis template or journal article.
Alternative justifications may focus on the research aim, sample relevance, scope of interpretation, and analytical depth.
The question of whether saturation is appropriate for different qualitative methodologies should be resolved through methodological consistency rather than convention.
Can a Qualitative Sample Size Be Calculated Using a Formal Model?
Specialized quantitative or probabilistic approaches to planning qualitative data collection have been proposed, particularly for objectives such as identifying a specified proportion of possible themes or estimating the likelihood of observing new information.
Such approaches can be useful under their stated assumptions.
However, they typically operationalize a particular aspect of qualitative adequacy rather than establish the sample needed for every interpretive objective.
For example, a model estimating the probability of identifying frequently occurring themes does not necessarily establish the number of participants required to develop a nuanced phenomenological interpretation.
Researchers should examine what the formal method actually estimates and whether that quantity corresponds to the study's purpose.
The existence of specialized calculations does not justify applying conventional statistical power formulas to qualitative interviews indiscriminately.
What About Grounded Theory and Theoretical Sampling?
Grounded theory commonly involves iterative data collection and analysis, with emerging concepts informing subsequent sampling decisions.
Theoretical sampling seeks information that can elaborate, compare, or challenge developing conceptual categories.
In traditions using theoretical saturation, recruitment may continue until categories are sufficiently developed for the intended theoretical account.
This approach cannot always be represented adequately by a fixed participant count established before analysis.
Nevertheless, researchers may need a provisional recruitment plan for ethics review and resource allocation.
The proposal should distinguish the initial planning target from the analytical criteria guiding subsequent recruitment.
What About Phenomenological Research?
Phenomenological research often emphasizes close examination of lived experience, although different phenomenological traditions vary in their analytical assumptions.
A relatively focused sample may be appropriate when participants have direct experience of the phenomenon and the analysis requires intensive engagement with individual accounts.
However, there is no single phenomenological sample-size formula applicable across all traditions.
The justification should explain why the participants and data support the intended phenomenological interpretation.
Simply citing a conventional number without discussing the analytical approach is insufficient.
What About Case Study Research?
Case study research introduces an additional distinction between the number of cases and the number of participants or sources within each case.
A study may investigate one institution using interviews, documents, observations, and administrative records.
In that situation, the adequacy of the evidence cannot be judged from interview count alone.
The researcher must explain the case boundaries, relevant units of analysis, and evidence needed to investigate the case adequately.
A statistical formula based on the number of employees in the institution would not ordinarily establish the number of interviews required for a qualitative case study.
How Should Mixed-Methods Studies Handle Sample Size?
Mixed-methods studies may require separate sample-size justifications for their quantitative and qualitative components.
For example, a sequential explanatory design might begin with a survey requiring a precision-based calculation and then conduct interviews to explore how participants interpret the survey findings.
The survey sample may be determined using statistical assumptions, while the interview sample is justified through purposive selection and the intended qualitative analysis.
The relationship between the two samples should also be explained, including how interview participants are selected from or connected to the quantitative phase.
One calculation should not automatically be used to justify both components.
What If a Thesis Committee Requires a Numerical Sample Size?
A requirement to specify a participant number does not necessarily require a statistical formula.
Researchers can provide a provisional target or range supported by the research aim, sampling criteria, analytical approach, comparable studies, and anticipated data richness.
For example, a proposal might identify a planned range of twelve to eighteen interviews and explain how that range was developed.
The numbers themselves are not universal recommendations. They become defensible only through the reasoning supporting their selection.
Where the methodology requires iterative sampling, the proposal should explain how recruitment decisions may change as analysis proceeds.
The issue of predetermining a qualitative thesis sample-size range should be discussed with reference to the chosen methodology and institutional requirements.
Should Population Size Determine Qualitative Interview Numbers?
Not ordinarily. A qualitative study involving teachers from a population of 10,000 does not necessarily require more interviews than a study involving teachers from a population of 500.
The appropriate sample depends on the phenomenon, relevant participant characteristics, variation required, and analytical objectives.
Population size may affect recruitment feasibility and the availability of relevant participant groups.
However, it does not automatically determine qualitative adequacy through the same relationship used in probability-sampling calculations.
Can Previous Studies Justify the Sample Without a Formula?
Previous studies can support a qualitative sample-size rationale when they are methodologically comparable.
Researchers should examine their aims, sampling strategies, populations, data collection methods, and analytical approaches.
For example, a published interview study involving a focused group of experienced teachers may inform planning for a similar investigation.
However, copying its participant count without examining the underlying rationale is not sufficient.
Prior studies provide contextual evidence, not automatic numerical authority.
How Can You Justify a Qualitative Sample Size in a Proposal?
A defensible justification should connect the recruitment plan to the study's purpose and methodological assumptions.
Begin by identifying the qualitative approach and explaining why the selected participants are capable of providing relevant information.
Describe the important variation the sample should capture and how the data will be analyzed.
Then establish an initial target or range, using relevant methodological literature and comparable research where appropriate.
Finally, explain how adequacy will be evaluated, including whether additional recruitment may be necessary.
The justification should make clear that the sample is planned to support qualitative interpretation, not statistical representativeness unless the study contains a separate quantitative component designed for that purpose.
What Should Be Reported After Data Collection?
The final research report should distinguish the planned sample from the achieved sample.
It should describe how participants were recruited, why they were eligible, and whether recruitment decisions changed during the study.
If saturation was used, the report should explain its definition and assessment. If information power or another framework informed the decision, the relevant reasoning should be documented.
Vasileiou and colleagues (2018) examined how qualitative interview sample-size sufficiency was characterized and justified in published research, highlighting the importance of transparent reporting.
The final participant count should be presented as an outcome of the study's sampling and analytical decisions, not as self-evident proof of adequacy.
