Should Quantitative and Qualitative Samples Overlap in a Convergent Mixed-Methods Design?

Should the People You Survey Also Be the People You Interview?

You are conducting a convergent mixed-methods study of students' online learning engagement. Your quantitative component includes a survey of several hundred students, while your qualitative component will involve interviews with a smaller group.

Should interview participants be selected from the survey respondents? Would using entirely different students weaken the study? And if the same participants contribute to both components, does that automatically make the findings easier to integrate?

These decisions become more complicated when the components investigate different stakeholder groups. For example, quantitative data may concern student performance while qualitative interviews involve teachers responsible for delivering the courses.

Sample overlap is not a universal requirement of convergent mixed methods. Its methodological value depends on the kind of relationship the researcher intends to examine between the quantitative and qualitative evidence.

Samples Should Overlap When Individual-Level Integration Requires It

Choosing and Justifying Sample Relationships in Convergent Mixed Methods

Why Does Sample Overlap Matter in a Convergent Design?

Convergent mixed-methods research brings quantitative and qualitative findings together to investigate a shared research problem.

Because the components are commonly collected during comparable periods and analyzed separately before integration, researchers must decide what relationships between participants or cases will support the intended comparison.

For example, a study may compare students' questionnaire responses about engagement with their interview accounts of learning activities.

If the same students contribute to both components and their records can be linked appropriately, the researcher can examine how each individual's quantitative responses correspond with their qualitative account.

If the components involve different students, the researcher may still compare broader patterns and themes, but cannot make the same direct individual-level comparisons.

Sample overlap therefore affects the scope of integration, not whether the study automatically qualifies as mixed methods.

What Sampling Arrangements Are Available?

Convergent studies can use several relationships between quantitative and qualitative samples.

These arrangements are not inherently ranked from strongest to weakest.

Their suitability depends on what the research question requires and whether the resulting evidence supports the intended inference.

When Is Using the Same Participants Most Appropriate?

Using the same participants is particularly useful when the study seeks to compare quantitative and qualitative evidence concerning individual experiences or characteristics.

Suppose researchers measure students' perceived usefulness of AI-generated feedback and interview those same students about how they interpret and apply feedback.

Individual-level linkage allows researchers to examine whether students with high usefulness ratings describe particular feedback practices, and whether students with similar ratings report different experiences.

This can reveal variation concealed by aggregate quantitative summaries.

However, the researcher must still consider whether the questionnaire and interview questions address comparable constructs.

Using the same participants does not make perceived usefulness and actual feedback application equivalent measures.

When Is a Nested Qualitative Sample Appropriate?

A nested sample is often useful when quantitative data are collected from a larger group and qualitative interviews are conducted with a purposively selected subset.

For example, a survey may involve 400 students, while interviews involve selected respondents representing relevant engagement experiences or demographic and contextual variation.

The qualitative sample can provide detailed evidence about selected cases while the quantitative sample supports its own statistical objectives.

However, the researcher should distinguish a convergent nested sample from an explanatory sequential sampling procedure.

In explanatory sequential research, completed quantitative findings typically guide subsequent qualitative selection.

In a convergent design, interview participants may be selected using criteria established before quantitative analysis, such as relevant experiences or participant characteristics.

If the researcher waits for completed quantitative results to determine whom to interview, the study may involve a sequential connection rather than purely concurrent collection.

The overall design should be described according to the actual methodological relationship, not the original label.

When Are Separate Samples Appropriate?

Separate samples may be appropriate when the study seeks to integrate broader quantitative patterns with qualitative experiences from the same defined population.

For example, a quantitative survey may estimate satisfaction among students across several courses, while qualitative interviews involve different students selected to represent relevant learning contexts.

The findings may be integrated to examine whether the qualitative accounts clarify the interpretation of the broader survey patterns.

However, the interviews cannot be linked to particular survey responses, and their themes should not automatically be interpreted as population prevalence estimates.

Separate samples may also reduce some forms of repeated-participation burden or influence.

The researcher should establish whether the populations and contexts are sufficiently comparable to support the intended interpretation.

Can Different Stakeholder Groups Be Used?

Yes. A convergent study may collect quantitative evidence from one stakeholder group and qualitative evidence from another when both contribute to a shared research problem.

For example, quantitative student performance data may be integrated with teacher interviews about implementation of an educational intervention.

Teachers' accounts could provide contextual evidence about instructional practices and implementation conditions.

However, those accounts cannot automatically explain why particular students obtained particular scores.

The integration may operate at the course, program, or institutional level, depending on how the data are linked.

Researchers should specify the relationship between stakeholder groups and avoid assuming that shared institutional membership establishes individual-level correspondence.

Does Sample Overlap Improve the Validity of Mixed-Methods Findings?

Not automatically. Overlap can strengthen the coherence of individual-level comparisons, but it does not independently establish the validity of either component.

A questionnaire may still have measurement limitations, while interviews may still be affected by sampling, questioning, or interpretive weaknesses.

Similarly, participants' retrospective accounts do not necessarily explain their quantitative scores.

Sample overlap can make certain comparisons possible, but the credibility of the integrated inference depends on the quality of the evidence and the appropriateness of the analytical relationship.

In some studies, separate samples may be preferable because they provide complementary perspectives or avoid unnecessary repeated participation.

How Does Overlap Affect the Unit of Analysis?

The unit of analysis determines what entity the findings describe and what relationships can be inferred.

When the same students contribute quantitative and qualitative data, integration may occur at the individual level.

When quantitative evidence concerns course averages and qualitative interviews concern selected students, integration may operate at the course or contextual level.

When different stakeholder groups are involved, the researcher must establish whether the findings can be related at an institutional or programmatic level.

The distinction is important because aggregate relationships do not necessarily describe individual cases.

Researchers should avoid ecological and individualistic inference errors when integrating findings from different analytical levels.

How Should Overlapping Participants Be Selected?

The qualitative sampling strategy should reflect the research question and the type of variation that needs to be investigated.

For example, researchers may purposively select students representing different academic programs, levels of experience, or patterns of platform use.

Palinkas and colleagues (2015) discuss purposeful sampling strategies in mixed-method implementation research, emphasizing the relationship between selection procedures and research objectives.

In a convergent design, the selection criteria may be established before quantitative analysis so that both components can proceed during comparable periods.

Researchers should avoid assuming that the qualitative sample must be statistically representative simply because participants also completed a survey.

Its adequacy depends on the qualitative question, relevant variation, and analytical approach.

Must the Quantitative and Qualitative Sample Sizes Be Equal?

No. Sample sizes should be justified independently according to the methodological purposes of the components.

Quantitative sample requirements may depend on precision, statistical power, model complexity, sampling design, or measurement objectives.

Qualitative sample adequacy depends on the research question, sampling strategy, depth of information, and analytical approach.

Malterud, Siersma, and Guassora (2016) discuss information power as one approach to considering qualitative interview sample adequacy.

There is no universal percentage of survey respondents who must also be interviewed.

For example, a quantitative sample of several hundred participants and a smaller purposive qualitative sample may be appropriate, provided each component supports its intended contribution.

Equal sample sizes are neither necessary nor sufficient for meaningful integration.

How Can Participants Be Linked Across Components?

When case-level integration is intended, researchers need an ethically appropriate method for associating quantitative and qualitative records with the same participant.

One approach is to assign pseudonymous study identifiers and maintain any necessary identity mapping separately from analytical datasets.

Researchers should specify who can access linkage information, how records will be stored, and how identifiable information will be managed.

Participants should be informed accurately about whether their responses can be linked across components.

Institutional ethics requirements and applicable data protection obligations should guide the procedures.

Linkage should be limited to what is necessary for the research purpose.

What If the Quantitative Survey Is Anonymous?

Genuinely anonymous survey responses ordinarily cannot be associated with identifiable interview participants after collection when no linkage mechanism exists.

A separate voluntary interview registration form may allow researchers to recruit participants, but it does not automatically establish which survey response belongs to each interviewee.

Researchers should distinguish anonymous data from confidential or pseudonymized data that retain a secure means of linkage.

If individual-level comparison is essential, the linkage procedure should be established before collection and accurately described during informed consent.

If the survey is genuinely anonymous, researchers may still integrate findings at a broader level, provided the resulting limitations are acknowledged.

Can Participation in One Component Influence the Other?

Potentially. When the same individuals complete a questionnaire and participate in interviews, the first activity may influence how they respond during the second.

For example, a questionnaire about academic integrity may prompt participants to reflect on practices they had not previously considered.

Their subsequent interview accounts may therefore differ from what they would have reported without completing the questionnaire.

Similarly, interviews may influence later survey responses by drawing attention to particular concepts.

These effects are not inevitable, but researchers should consider collection order and participant burden.

Possible safeguards include documenting the order of participation, using consistent procedures, and selecting an appropriate interval between activities when justified.

What Happens When Some Participants Complete Only One Component?

Partial participation may create a partially overlapping sample even when complete overlap was originally intended.

For example, some survey respondents may decline interviews, while others may participate in interviews but fail to complete the questionnaire.

The researcher should distinguish participants with linked records from those contributing only one form of evidence.

Case-level integration should be limited to participants for whom appropriate linkage exists.

Other participants may still contribute to component-specific analyses or broader contextual interpretation.

Researchers should examine whether nonparticipation systematically affects relevant characteristics and whether missing profiles limit the integrated conclusions.

How Do You Integrate Findings From Overlapping Samples?

Overlapping samples can support case-oriented integration through joint displays or structured comparisons.

For example, a joint display might present selected participants' quantitative engagement scores, qualitative accounts of learning activities, and an integrated interpretation.

Guetterman, Fetters, and Creswell (2015) discuss joint displays as a means of facilitating integration of quantitative and qualitative findings.

However, the display should not expose identifiable participant information, particularly when unusual combinations of characteristics could permit reidentification.

The researcher must also explain what the comparison establishes rather than simply presenting numerical scores beside quotations.

How Do You Integrate Findings From Separate Samples?

When samples do not overlap, integration generally requires a defensible conceptual, contextual, or group-level relationship.

For example, survey findings concerning students' overall satisfaction may be considered alongside interviews with different students about particular platform experiences.

The qualitative accounts may clarify possible interpretations of the survey findings without directly explaining particular respondents' scores.

Researchers should examine whether the samples represent sufficiently related populations and conditions.

They should also avoid treating qualitative themes as estimates of prevalence or assuming that aggregate quantitative findings apply uniformly to interview participants.

Separate samples can support meaningful convergence, but the integrated conclusions must reflect the available linkage.

Does Sample Overlap Guarantee Convergence?

No. The same participants may provide quantitative and qualitative evidence that appears inconsistent.

For example, a student may report high satisfaction on a questionnaire but describe substantial frustration during an interview.

These findings may reflect differences between overall evaluation and specific experiences, changes in interpretation, response context, or genuine disagreement.

Researchers should examine the constructs and circumstances before treating the findings as contradictory.

Convergence concerns the relationship between findings, not the identity of the participants who produced them.

How Should the Sampling Relationship Be Reported?

The methodology should specify whether the samples were identical, nested, partially overlapping, or separate.

Researchers should explain the sampling rationale for each component, any record-linkage procedures, and the level at which integration was intended.

Where overlap was planned but incomplete, the report should describe recruitment outcomes and their implications.

When different stakeholder groups or populations are involved, the researcher should establish how their evidence relates to the overarching question.

The integrated interpretation should not imply stronger individual-level connections than the sampling arrangement supports.

Choosing Between Overlapping and Separate Samples in Online Learning Research

Suppose the quantitative results indicate that frequent platform access is associated with favorable engagement ratings, while interviews reveal that students use the platform for different purposes.

Under Option A, the researcher can examine whether students with particular quantitative profiles describe different learning practices.

Under Option B, the researcher can examine how qualitative accounts qualify the broader interpretation of platform activity, but cannot attribute those accounts to particular quantitative respondents.

Neither arrangement automatically establishes causal relationships or population prevalence of the qualitative themes.

The appropriate choice depends on the intended integration and the conclusions the research question requires.

Common Misconceptions About Sample Overlap in Convergent Research

How to Choose the Right Sampling Relationship for Your Convergent Study

Begin by identifying the level at which the quantitative and qualitative findings must be related.

If the research question requires comparison of individual scores and experiences, overlapping participants may be particularly useful. If the purpose is broader contextual interpretation, separate samples may be sufficient.

Document the sampling rationale before collection, including how participants will be selected, whether records will be linked, and what comparisons the final evidence can support.

The appropriate sampling arrangement is the one that serves the overarching research question while preserving methodological rigor and participant protections.

Check Whether Your Convergent Samples Should Overlap

Frequently Asked Questions About Overlapping Samples in Convergent Mixed Methods

Choose Sample Overlap According to the Integration You Need

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.