How Does Sample Specificity Affect Qualitative Sample Size?

Can Recruiting More Relevant Participants Reduce the Number of Interviews You Need?

You want to investigate how university teachers make decisions about using generative artificial intelligence when designing performance-based assessments. Your institution has hundreds of teachers, but not all of them have developed online courses, designed performance-based tasks, or used AI during assessment development.

Should you recruit broadly from the entire teaching population, or focus on teachers with direct experience of the phenomenon? Would a highly specific sample allow you to conduct fewer interviews without weakening the study?

These questions concern sample specificity, an important consideration in qualitative research. The issue is not simply how many people participate, but how closely their experiences align with the research aim and what kinds of information their participation can provide.

Greater Sample Specificity May Increase Information Power

How Sample Specificity Influences Qualitative Recruitment and Adequacy

What Is Sample Specificity in Qualitative Research?

Sample specificity refers to the extent to which selected participants possess characteristics, experiences, or knowledge directly relevant to the phenomenon under investigation.

For example, a study examining teachers' decisions about AI-assisted assessment design may benefit from recruiting teachers who have actually developed assessments using or evaluating generative AI.

By contrast, recruiting teachers who have only heard about AI in education may produce information about general attitudes rather than the assessment-design decisions the researcher intends to understand.

Specificity is therefore relational. A participant is not inherently information-rich in every study. Their relevance depends on the particular research question.

How Does Sample Specificity Relate to Information Power?

Malterud, Siersma, and Guassora (2016) identify sample specificity as one of five dimensions of information power, alongside study aim, established theory, dialogue quality, and analysis strategy.

Their framework proposes that samples containing information closely aligned with the research aim may require fewer participants than samples offering less specific information.

However, specificity operates alongside the other dimensions. A highly relevant participant may still provide a limited account, and a narrowly recruited sample may omit experiences essential to the analysis.

The information power framework therefore supports a conditional relationship between sample specificity and participant numbers, not a numerical formula.

Is Sample Specificity the Same as Homogeneity?

No. Specificity concerns relevance to the research aim, while homogeneity concerns similarity among participants on particular characteristics.

A sample can be highly specific without being homogeneous in every respect.

For example, all participants might have experience designing AI-assisted assessments, while differing in discipline, teaching experience, assessment type, or institutional setting.

These differences may be valuable if the research seeks to understand how assessment-design decisions vary across contexts.

Conversely, a homogeneous group of teachers may be poorly suited to the study if few have experienced the phenomenon being investigated.

How Does the Research Aim Determine Relevant Participant Characteristics?

Participant characteristics should be selected according to what the research question requires.

Consider three different studies involving generative AI in higher education.

Although all three studies concern educational AI, their participant populations differ because their research aims differ.

The relationship between the research aim and qualitative sample size should therefore be established before eligibility criteria and recruitment targets are finalized.

How Should Inclusion Criteria Be Developed?

Inclusion criteria should identify participants who can provide relevant information about the phenomenon.

For example, a study of online assessment designers might require participants to have developed at least one performance-based assessment for a fully online course.

If the study specifically concerns decisions involving generative AI, participants may also need experience considering, adopting, adapting, or rejecting AI for assessment-development purposes.

Importantly, the criteria should not require participants to hold favorable opinions or report successful experiences. Teachers who deliberately rejected AI may provide valuable contrasting accounts.

Eligibility criteria should be sufficiently specific to support the aim without unnecessarily excluding experiences that could challenge emerging interpretations.

Can Eligibility Criteria Become Too Restrictive?

Yes. Excessively narrow criteria can reduce recruitment feasibility and exclude relevant variation.

Suppose a researcher requires participants to have at least ten years of teaching experience, hold a doctoral degree, teach a particular discipline, and have implemented five AI-assisted assessments.

Unless each condition is necessary for the research aim, the resulting sample may be narrower than the phenomenon requires.

The study could then produce detailed accounts from a very specialized group while overlooking experiences that matter to the intended interpretation.

Researchers should justify each eligibility criterion substantively rather than assume that more restrictions automatically increase sample quality.

How Does Purposive Sampling Support Specificity?

Purposive sampling involves selecting participants because their characteristics or experiences are relevant to the research purpose.

It is particularly useful when not every member of the broader population has encountered the phenomenon.

For example, recruiting teachers through a general university mailing list may generate many volunteers with limited assessment-design experience. Targeted recruitment through course-design responsibilities may identify more directly relevant participants.

However, purposive sampling is not automatically rigorous simply because it is labeled purposive. The researcher must explain the selection criteria and how the resulting sample supports the intended analysis.

What Is Criterion Sampling?

Criterion sampling is a purposive strategy in which participants must meet specified conditions relevant to the research question.

For example, participants might be required to have designed and implemented an online assessment during the previous academic year.

The criterion establishes direct experience of the activity being investigated.

However, criterion sampling does not determine the number of participants required. The sample must still provide adequate information for the intended analytical scope.

Criteria should be transparent, defensible, and consistent with the phenomenon being studied.

What Is Maximum Variation Sampling?

Maximum variation sampling seeks participants who differ across characteristics considered important to understanding the phenomenon.

For example, a researcher might recruit teachers from different disciplines, levels of teaching experience, or institutional settings while requiring all participants to have direct experience with AI-assisted assessment design.

This strategy combines specificity concerning the phenomenon with variation concerning relevant contexts.

It can help reveal similarities and differences that a narrowly homogeneous sample might miss.

However, greater variation may also increase the amount of information needed to support meaningful cross-case interpretation.

Does Greater Participant Diversity Always Require More Interviews?

Not necessarily, but diversity can increase sample-size requirements when the research aim involves understanding differences among participant groups.

If the study seeks to compare teachers across several disciplines, it needs sufficient information within those contexts to support the comparison.

A sample containing one participant from each of ten disciplines may offer breadth but insufficient depth for claims about disciplinary differences.

Conversely, a study seeking an overall interpretive account may not require extensive subgroup comparisons.

The required variation depends on the analytical claims, not on an obligation to represent every possible demographic category.

Can a Highly Specific Sample Be Too Small?

Yes. Participant relevance does not eliminate the need for sufficient information.

A sample of three highly experienced teachers might provide detailed accounts, but it may not capture important differences in assessment practices or institutional expectations.

Whether such a sample is adequate depends on the methodology, research aim, and depth of analysis.

For an intensive case-oriented investigation, a small number of participants may be defensible. For a broader comparative study, it may be insufficient.

Sample specificity increases the potential relevance of participants, but it does not guarantee analytical completeness.

Does Recruiting Experts Always Increase Information Power?

No. Expertise is useful only when it is relevant to the research question.

A senior academic administrator may possess extensive knowledge of institutional policy but limited firsthand experience developing classroom assessments.

For a study of policy implementation, that participant may be highly relevant. For a study of teachers' assessment-design decisions, a classroom practitioner may provide more directly useful information.

Researchers should avoid equating professional seniority, academic qualifications, or institutional authority with informational relevance.

Experience of the phenomenon is often more important than status.

Can Typical Participants Be More Informative Than Exceptional Participants?

Yes, depending on the research aim.

If the objective is to understand routine assessment-design practices, participants whose experiences reflect common institutional conditions may be particularly relevant.

If the objective is to investigate unusual innovations or exceptional challenges, atypical cases may be more informative.

Neither typical-case nor extreme-case sampling is inherently superior.

The sampling strategy should match the kind of understanding the researcher seeks to develop.

How Do Negative and Discrepant Cases Affect Sample Specificity?

Negative or discrepant cases are participants whose experiences challenge an emerging interpretation.

Suppose most teachers report that generative AI reduces assessment-development time, while others describe additional work caused by verification and revision.

The contrasting experiences may reveal conditions under which AI changes workload differently.

Recruiting such participants can strengthen analysis by examining the limits of emerging patterns.

A highly specific sample should not become a collection of participants selected because they confirm the researcher's expectations.

Does Snowball Sampling Improve Specificity?

Snowball sampling may help researchers identify participants with relevant experiences that are difficult to locate through ordinary recruitment channels.

For example, experienced course designers may recommend colleagues who have encountered unusual assessment-integrity problems.

However, referrals can concentrate recruitment within similar professional networks.

This may reduce variation and exclude participants whose experiences differ from those of the initial contacts.

Researchers should therefore assess whether referral-based recruitment supports the intended sampling logic and consider additional recruitment channels where necessary.

How Does Sample Specificity Affect Data Saturation?

A relatively specific participant group may produce recurring information sooner than a heterogeneous group, particularly when the research aim is focused.

However, this relationship is conditional and depends on what saturation means in the study.

Code repetition does not necessarily establish that meanings, contextual variation, or conceptual relationships have been sufficiently explored.

Moreover, saturation is not appropriate for every qualitative methodology.

Researchers should avoid claiming that a highly specific sample automatically reaches saturation after a small number of interviews.

How Does Sample Specificity Affect Recruitment Feasibility?

Highly specific eligibility criteria may reduce the number of accessible participants.

For example, a university may employ hundreds of teachers but only a small subset may have experience designing AI-assisted performance-based assessments.

The researcher should assess whether the eligible population is sufficiently large and accessible for the proposed study.

Recruitment feasibility does not establish methodological adequacy. A small eligible population may justify investigating all accessible cases, but it does not automatically support broad conclusions.

If recruitment is constrained, the researcher may need to reconsider the scope of the aim or the sampling strategy while preserving scientific coherence.

Should Sample Specificity Be Reviewed During Data Collection?

Yes, where the methodology permits iterative recruitment.

Initial interviews may reveal that the original eligibility criteria identify participants with relevant experience but overlook a consequential perspective.

For example, the study may initially recruit teachers who actively use generative AI, then discover that teachers who deliberately avoid it can illuminate important professional decisions.

The researcher may revise recruitment to include these experiences, subject to the study protocol and relevant ethics requirements.

Such changes should be documented rather than presented retrospectively as part of an unchanged sampling plan.

How Should Sample Specificity Be Reported?

The methods section should explain why the selected participants are particularly relevant to the phenomenon.

Researchers should report the eligibility criteria, purposive sampling strategy, important participant characteristics, and rationale for including relevant variation.

They should also explain how the achieved sample supported the intended analysis.

A statement that participants were selected purposively is insufficient without describing what made them appropriate for the study.

Likewise, describing the sample as highly specific does not replace an assessment of data quality or analytical adequacy.

Recruiting Teachers With Direct Experience of AI-Assisted Assessment Design

The revised sampling strategy may improve information power because participants have experiences closely aligned with the research question.

However, the final sample size cannot be established from specificity alone. The researcher must still consider interview quality, meaningful variation, and the requirements of the chosen analytical approach.

Common Misconceptions About Sample Specificity

Recruit for Relevance Without Excluding Necessary Variation

Begin with the research aim and identify the experiences participants must possess to contribute meaningfully.

Then distinguish essential eligibility criteria from characteristics that should vary across the sample.

Sample specificity should improve the fit between the research question and the people recruited.

The final sample-size justification should explain both why participants were selected and why their collective accounts were sufficient for the intended analysis.

Before Finalizing a Specific Qualitative Sample

Frequently Asked Questions About Sample Specificity

Sample Specificity Improves Relevance, Not Automatically Adequacy

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.