03 · What You Need to Know
How Participant Overlap Affects Sequential Mixed-Methods Research
Why Does Participant Overlap Matter?
Sequential mixed-methods research involves a relationship between an initial component and a subsequent component. The sampling arrangement determines which observations, experiences, or findings can be connected across those phases.
For example, a researcher may identify students with unusually high platform engagement but relatively low academic performance. Interviewing those same students allows the researcher to examine how their accounts relate to their measured profiles.
By contrast, interviewing different students may provide useful contextual understanding but cannot directly explain the experiences of the particular students whose quantitative results were identified.
Participant overlap therefore affects the level of integration that is possible.
However, direct individual-level linkage is not necessary for every sequential design. A qualitative phase that informs questionnaire development may be followed by quantitative testing with an entirely different sample.
The appropriate choice depends on the specific integration problem the study is intended to address.
What Sampling Arrangements Are Possible Across Phases?
Sequential studies can use several sampling relationships. These arrangements differ in the degree of participant overlap and the kinds of integration they support.
| Sampling arrangement |
Relationship between phases |
Potential use |
| Identical participants |
All participants contribute to both components. |
Directly examine quantitative and qualitative evidence concerning the same individuals. |
| Nested or overlapping sample |
A subset of first-phase participants contributes to the second phase. |
Investigate selected quantitative profiles through qualitative follow-up. |
| Different participants from the same population |
Both phases recruit from a common target population without individual overlap. |
Develop instruments or investigate related population-level phenomena. |
| Different but related populations |
Phases involve different stakeholder groups or analytically related populations. |
Examine contextual processes or develop measures for a different target group. |
| Partially overlapping samples |
Some participants contribute to both phases while others contribute to only one. |
Combine individual-level connections with broader or additional perspectives. |
These arrangements are not automatically interchangeable. Each requires a justification that reflects the study's questions and the intended relationship between findings.
When Is Using the Same Participants Particularly Useful?
Participant overlap is especially useful when the researcher intends to investigate particular quantitative results at the individual level.
Suppose a survey identifies teachers who report high confidence in using generative AI but low implementation frequency. Interviewing selected teachers from that group allows the researcher to investigate how those individuals understand confidence and make implementation decisions.
Similarly, students with contrasting engagement and performance profiles can be selected for interviews about their learning practices.
These arrangements are common in explanatory sequential mixed-methods research.
However, using the same participants is valuable only when their individual-level information contributes to the explanatory question. It should not be treated as a universal requirement.
When Is a Subsample of the Quantitative Participants Appropriate?
Many explanatory sequential studies use a qualitative subsample drawn from the original quantitative participants.
The quantitative phase may involve hundreds of respondents, while the qualitative phase focuses on a smaller number of information-rich cases.
For example, researchers might select students representing contrasting combinations of platform participation and academic performance.
The purpose is not to reproduce the quantitative sample statistically. It is to investigate selected patterns in sufficient depth.
Purposive selection may involve typical cases, extreme cases, contrasting profiles, relevant subgroups, or cases that differ from statistical expectations.
Palinkas and colleagues (2015) discuss purposeful sampling strategies in mixed-method implementation research, emphasizing the importance of aligning selection with the research purpose.
The decision about which quantitative participants should be interviewed should therefore follow the explanatory question rather than a fixed sampling percentage.
When Are Different Participants More Appropriate?
Different participants may be appropriate when the first phase is intended to inform the development of procedures that will subsequently be evaluated with another sample.
For example, researchers might interview teachers to understand how they conceptualize responsible AI-supported assessment. Those findings could inform a questionnaire administered to a larger sample of teachers.
The qualitative participants provide evidence for construct development, while the quantitative participants provide evidence for measurement evaluation or broader investigation.
This arrangement is common in exploratory sequential mixed-methods research.
Using different participants may also help assess whether a proposed instrument performs adequately beyond the individuals who contributed to its development.
However, a new sample does not automatically establish generalizability. The quantitative sampling design and intended population remain important.
Must Exploratory Sequential Research Use Different Samples?
No. The same participants may contribute to qualitative exploration and subsequent quantitative investigation when that arrangement supports the research purpose.
For example, a researcher might interview members of a small professional community and later invite them to evaluate an instrument developed from the qualitative findings.
However, using the same individuals for both development and evaluation can introduce limitations.
Participants may recognize questionnaire content derived from their earlier interviews, and their familiarity with the development process may influence their responses.
When independent evaluation is important, a separate quantitative sample may be preferable.
Researchers should distinguish preliminary testing with development participants from broader evaluation of the instrument's measurement properties.
Can Different Stakeholder Groups Be Used Across Phases?
Yes, provided the relationship between the stakeholder groups supports the overarching research question.
For example, a quantitative phase might examine student outcomes in an educational intervention, while a qualitative phase interviews teachers responsible for implementing the intervention.
Teachers' accounts could provide evidence about implementation conditions relevant to interpreting the student outcomes.
However, the researcher should not claim that teachers' accounts directly explain the experiences of particular students unless appropriate supporting evidence exists.
The integrated interpretation may operate at the classroom, program, or institutional level rather than the individual student level.
Different stakeholder samples can provide complementary perspectives, but their relationship must be established rather than assumed.
What Is the Difference Between Individual-Level and Group-Level Integration?
Individual-level integration connects quantitative and qualitative evidence concerning identifiable corresponding cases.
For example, a student's quantitative engagement profile may be linked with that same student's interview account.
Group-level integration examines relationships between findings aggregated or interpreted at a broader level, such as classrooms, institutions, or participant categories.
Individual-level integration
Connects different forms of evidence concerning the same participant or case, subject to appropriate consent and data governance.
Group-level integration
Relates findings concerning groups, institutions, contexts, or populations without necessarily linking individual records.
The distinction matters because findings at one level do not automatically establish relationships at another.
For example, a university with high average student engagement may also have teachers reporting favorable implementation experiences. That institutional-level correspondence does not establish that particular engaged students were taught by particular teachers or experienced the same implementation conditions.
Researchers should avoid ecological and individualistic inference errors when integrating evidence across analytical levels.
Does Using the Same Participants Improve Validity?
Not automatically. Participant overlap can strengthen the coherence of individual-level integration, but it does not independently establish the validity of either component.
A quantitative instrument may still have measurement limitations, while interviews may still be affected by sampling, questioning, or interpretive weaknesses.
Likewise, participants' accounts do not necessarily provide definitive explanations for their quantitative outcomes.
Using the same participants can support more direct comparison, but the quality of the resulting inference depends on the appropriateness of the measures, qualitative procedures, analytical approach, and integration.
In some circumstances, different samples may be more appropriate because they provide independent evaluation or relevant perspectives unavailable from the original participants.
How Does Participant Attrition Affect Sequential Research?
Participant attrition occurs when individuals who contributed to an earlier phase do not participate in a later phase.
In sequential mixed-methods research, attrition can be consequential when the second phase depends on recruiting particular first-phase participants.
For example, students representing unusual quantitative profiles may be difficult to contact or unwilling to participate in interviews.
If the researcher replaces them with participants who do not share the relevant profiles, the intended explanatory connection may be weakened.
Researchers should plan recruitment procedures, identify eligible alternatives, and document how nonparticipation affects the final sample.
They should also consider whether those who agree to follow-up differ systematically from those who decline.
Such differences may limit the range of experiences represented in the qualitative findings.
What If the First-Phase Data Were Collected Anonymously?
Genuinely anonymous quantitative data cannot ordinarily be linked back to identifiable respondents for individual follow-up.
If a survey does not collect identifying information or a permissible linkage mechanism, researchers cannot reliably determine which individual produced a particular response profile.
A separate voluntary contact form may allow recruitment for interviews, but it does not automatically enable linkage between interviewees and their original survey responses.
Researchers should distinguish anonymous data from confidential or pseudonymized data that retain a secure means of linkage.
Where individual-level integration is necessary, the sampling and data governance arrangements should be established before data collection and reviewed under applicable ethical requirements.
Watch Out
Do not promise participants that their survey responses are anonymous if the research team retains a mechanism that can link those responses to their identities. Explain the actual confidentiality and linkage arrangements accurately during informed consent.
How Can Researchers Link Participants Across Phases Ethically?
When individual-level linkage is required, researchers may use securely managed participant identifiers or another ethically approved procedure.
The linkage mechanism should be proportionate to the research purpose and should protect participants from unnecessary disclosure.
Researchers should consider who can access the linkage information, how records will be stored, how long identifiable information will be retained, and whether participants have consented to follow-up contact and data linkage.
These arrangements may be particularly important when quantitative profiles concern sensitive characteristics, academic performance, or potentially identifiable small subgroups.
Institutional ethics requirements and applicable data protection rules should guide the procedures.
The broader methodological question of connecting samples across sequential phases includes both the analytical purpose and the practical safeguards required for linkage.
How Many Participants Should Overlap Across Phases?
There is no universal percentage of participants who must contribute to both phases.
In an explanatory sequential study, the number of interview participants should be justified according to the qualitative research question, sampling strategy, relevant variation, and analytical approach.
In exploratory sequential research, the qualitative and quantitative samples may differ substantially because they serve different developmental and inferential purposes.
For example, interviews with 18 teachers might inform questionnaire development, while a subsequent quantitative evaluation involves several hundred teachers.
The numerical difference does not establish that the design is weak or that one phase is necessarily less important.
Malterud, Siersma, and Guassora (2016) discuss information power as one approach to qualitative sample adequacy, emphasizing the relationship between sample needs and the study's purpose and analytical characteristics.
Quantitative sample requirements should be evaluated separately according to the intended statistical analyses and measurement objectives.
Can the Same Participants Create Bias Across Phases?
Potentially. Repeated participation can affect how individuals respond, particularly when earlier research activities influence their awareness, expectations, or subsequent behavior.
For example, completing a detailed questionnaire about AI-supported assessment may prompt teachers to reflect on practices they had not previously considered. Their later interview accounts may therefore be influenced by the survey experience.
Similarly, participants involved in developing an instrument may respond differently during its subsequent evaluation because they recognize the concepts or wording.
These effects are not inevitable and do not automatically invalidate overlapping samples.
Researchers should consider whether repeated participation could influence the phenomenon under investigation and whether the resulting limitations are acceptable for the intended research purpose.
How Should the Sampling Relationship Be Reported?
The methodology should state whether the phases used identical, nested, partially overlapping, or different samples.
Researchers should explain how participants were selected, whether quantitative findings informed recruitment, and what relationships between individual or group-level evidence were possible.
For overlapping samples, the report should describe the linkage procedure without disclosing information that could compromise confidentiality.
For different samples, the researcher should explain how the populations or contexts are related and why that relationship supports integration.
When participants decline follow-up, the recruitment outcomes and implications should be reported where relevant.
The final interpretation should not imply a stronger connection between phases than the sampling arrangement actually supports.
