Should Quantitative Results Determine Which Participants to Interview Next?

Should Your Survey Results Determine Who You Interview?

You have completed a quantitative survey and identified several patterns that warrant further investigation. Some participants reported unusually high engagement, others reported low engagement, and a few demonstrated results that were inconsistent with your expectations.

You now want to conduct interviews. Should you randomly select respondents, invite everyone who volunteered, or deliberately recruit participants whose quantitative results represent particular patterns?

This decision matters because the participants selected for qualitative follow-up influence what the second phase can explain. Interviewing readily available respondents may produce valuable accounts, but those accounts may not address the quantitative findings that motivated the follow-up.

The methodological challenge is to select participants because they can illuminate a specific research problem, without treating their numerical scores as complete descriptions of who they are or what they have experienced.

Use Quantitative Results to Guide Interviews When Specific Cases Can Clarify the Findings

How Quantitative Findings Can Inform Qualitative Participant Selection

What Does Connecting Samples Mean in Mixed Methods?

Connecting samples means using information from one component of a mixed-methods study to inform participant selection for another component.

In a quantitative-first study, researchers may use survey responses, measured outcomes, statistical profiles, or other quantitative findings to identify participants for qualitative follow-up.

Fetters, Curry, and Creswell (2013) identify connecting as an important methods-level integration approach. The relationship between the components is established through sampling decisions rather than only through the final comparison of findings.

For example, a researcher might identify students whose measured engagement is high but whose assessment performance is low, then invite selected students to discuss their learning experiences.

The qualitative component can investigate how participants understand their engagement, what activities they undertake, and which contextual circumstances may help interpret the quantitative pattern.

This is particularly useful in an explanatory sequential mixed-methods design, where initial quantitative findings inform a subsequent qualitative investigation.

Why Not Simply Interview a Random Selection of Survey Respondents?

Random sampling may be appropriate for certain research purposes, but it does not necessarily produce the most informative qualitative sample for explaining particular quantitative findings.

Suppose a survey identifies a relatively small group of teachers who report high confidence in using generative AI but low implementation frequency.

A random selection of respondents might include few teachers from this group. Consequently, the qualitative phase may not adequately investigate the discrepancy that motivated the follow-up.

Purposive sampling allows researchers to select cases according to their relevance to the explanatory question.

However, random selection can still be useful within a defined quantitative subgroup when there are more eligible participants than the researcher can interview. The sampling strategy should follow the inferential purpose rather than a blanket preference for either random or purposive selection.

Which Quantitative Patterns Can Guide Interview Selection?

Several sampling strategies may be appropriate, depending on what the researcher wants to understand.

These approaches can overlap. A researcher might select contrasting cases from several subgroups or use maximum variation sampling within a defined quantitative profile.

The selection criteria should be established in relation to the research question and applied transparently.

Should You Select Participants With the Highest and Lowest Scores?

Not automatically. Extreme case sampling can be informative when unusually high or low values are central to the research question. However, selecting the highest and lowest scores simply because they are easy to identify may produce an unhelpful or distorted qualitative sample.

Extreme values can reflect measurement error, temporary circumstances, or random variation. Researchers should consider the reliability of the underlying measures and the substantive meaning of the selected thresholds.

For example, students with exceptionally low recorded platform activity may have engaged in substantial offline learning. Interviews could help investigate this possibility, but the researcher should not assume that low activity necessarily means low engagement.

Likewise, selecting only extreme cases may overlook the experiences of participants whose results fall within the more common range.

When the objective is to understand variation across the distribution, a broader purposive sampling strategy may be more appropriate.

How Do You Select Participants With Unexpected Results?

Unexpected case selection involves identifying participants whose quantitative outcomes differ meaningfully from a theoretical expectation, a prespecified comparison, or an appropriately estimated statistical model.

Suppose a researcher finds a positive association between recorded online participation and academic performance. Some students may nevertheless demonstrate relatively high participation and low performance, or low participation and high performance.

These cases could be informative because they challenge a simple interpretation of the overall association.

However, researchers should be cautious when defining unexpected cases using statistical residuals. Residual values depend on model specification, measurement quality, and the variables included in the analysis.

An unusual residual does not establish that a participant is genuinely anomalous or that an unmeasured explanatory factor must exist.

Qualitative follow-up should investigate possible interpretations while acknowledging the uncertainty associated with the original quantitative classification.

Should Interview Participants Come From the Original Quantitative Sample?

Using participants from the original quantitative sample can be advantageous when the researcher intends to connect individual quantitative profiles with qualitative accounts.

For example, interviewing students whose survey responses are already available allows the researcher to examine how each participant's account relates to their measured characteristics.

This can support case-level integration, provided that participants have consented to the relevant data linkage and appropriate confidentiality safeguards are maintained.

However, the same participants are not always required. Researchers may recruit additional participants representing relevant categories or contexts when the explanatory purpose does not require direct individual-level linkage.

Different sampling arrangements support different kinds of inference. Researchers should not imply that qualitative accounts explain particular individuals' quantitative outcomes when the datasets cannot actually be connected at that level.

The broader issue of whether the same participants should be used across phases depends on the design's integration requirements.

How Many Participants Should Be Selected for Interviews?

There is no universal interview sample size or fixed percentage of the quantitative sample that guarantees an adequate qualitative follow-up.

The number of participants depends on the explanatory question, sampling strategy, diversity of cases, analytical approach, and the depth of information required.

Malterud, Siersma, and Guassora (2016) propose information power as one way of thinking about qualitative sample adequacy. Under this approach, sample needs depend partly on the study aim, sample specificity, theoretical background, quality of dialogue, and analytical strategy.

This does not provide a universal numerical rule. Different qualitative traditions also approach sample adequacy differently.

For example, a study examining a narrowly defined group of participants with a highly specific quantitative profile may require a different sampling strategy from a study comparing experiences across several heterogeneous subgroups.

Researchers should justify their qualitative sample in relation to what the interviews are intended to explain, not merely the number of survey respondents available.

Should Participants Be Selected Before Quantitative Analysis Is Complete?

When the qualitative phase is intended to investigate particular quantitative findings, participant selection generally should follow sufficient quantitative analysis to identify those findings reliably.

Recruitment preparation can begin earlier. Researchers may obtain permission to contact respondents for follow-up, establish secure linkage procedures, and identify potential recruitment constraints.

However, final selection should reflect the results that the qualitative phase is supposed to investigate.

Premature selection can undermine the intended explanatory connection if later analysis reveals different patterns or changes the classification of relevant cases.

The decision about which first-phase findings should guide the second phase should therefore be made explicitly.

How Should Quantitative Selection Criteria Be Defined?

Selection criteria should be transparent, substantively meaningful, and appropriate to the quantitative evidence.

Researchers may use theoretically justified thresholds, relevant combinations of variables, subgroup membership, or statistical profiles.

For example, a study could identify participants reporting high perceived competence but low implementation frequency. The researcher should explain how high and low were defined, why those categories matter, and how measurement uncertainty was considered.

Arbitrary median splits or other data reductions may discard information and create misleading distinctions. When continuous variables are categorized for sampling, the researcher should justify the decision and retain the original values for interpretation where appropriate.

Selection criteria should also consider whether participants are eligible, available, and willing to take part in the qualitative phase.

What If Selected Participants Decline the Interview?

Nonparticipation is a practical and methodological concern, particularly when the quantitative profile of interest is uncommon.

Researchers should anticipate refusals and establish a recruitment plan that preserves the intended sampling logic.

For example, they might identify additional eligible participants within each profile or use a transparent procedure for inviting replacements.

However, replacing an unavailable participant with someone from a substantially different profile may weaken the explanatory purpose.

Researchers should document recruitment outcomes and consider whether nonparticipation may have affected the range of experiences represented in the qualitative findings.

How Do You Protect Participants When Linking Survey and Interview Data?

Connecting quantitative and qualitative records may increase the risk of identifying participants, particularly when the sample includes small subgroups or unusual profiles.

Researchers should establish appropriate consent procedures, secure linkage methods, access controls, and data minimization practices.

If the original quantitative survey was genuinely anonymous, researchers may be unable to identify particular respondents for follow-up. A separate voluntary contact process may permit recruitment, but it does not automatically enable individual-level linkage.

Researchers should not promise anonymity while retaining information that allows responses to be linked to identifiable individuals.

Ethical requirements vary by institution and jurisdiction. The recruitment and linkage procedures should comply with applicable research ethics approval and data protection obligations.

How Should Interview Findings Be Integrated With Quantitative Profiles?

Integration should go beyond reporting interview themes from participants selected using quantitative results.

Researchers should examine how the qualitative accounts relate to the quantitative patterns that informed selection.

A joint display may organize selected profiles, corresponding qualitative findings, and integrated interpretations.

For example, a researcher might compare the experiences of students with high recorded engagement and low assessment performance against those with similar engagement but higher performance.

The analysis could identify differences in reported study practices, interpretations of learning activities, or contextual constraints.

However, such differences should not automatically be treated as causal explanations. The purposive sample is designed to investigate experiences and possible processes, not to estimate the prevalence or effects of those processes in the original population.

Researchers should explain how samples are connected across phases and what level of integrated inference the linkage supports.

Can Selection Based on Quantitative Results Introduce Bias?

Yes. Selecting participants according to their quantitative outcomes can create interpretive risks if researchers overlook how the selection process shapes the qualitative evidence.

For example, interviewing only high-performing participants may produce accounts that are not representative of lower-performing participants' experiences. Selecting only respondents who support a preferred hypothesis may encourage confirmation bias.

Selection on extreme observed values can also be affected by measurement error and regression toward the mean, particularly when outcomes are measured repeatedly.

These concerns do not make purposive selection inappropriate. They mean that the resulting qualitative evidence should be interpreted in relation to the selection criteria.

Selecting Interview Participants From an Online Learning Survey

Suppose interviews suggest that some students access the platform frequently for administrative purposes, while others engage in substantial offline preparation that is not captured by platform records.

These findings could indicate that recorded participation represents different kinds of activity across students and may not fully capture meaningful learning engagement.

The integrated interpretation would be stronger than simply reporting that participation and performance are associated. Nevertheless, the interviews would not establish that particular study strategies caused the observed performance differences.

The purpose of the qualitative sample is to investigate selected patterns, not to reproduce the statistical representativeness of the original quantitative sample.

Common Mistakes When Selecting Interview Participants From Quantitative Results

How to Choose an Interview Sampling Strategy That Fits Your Findings

Start with the quantitative results that require explanation. Identify which participants or cases could provide information relevant to the unresolved question.

Then select a sampling strategy that reflects the kind of explanation sought. Do not begin by choosing a convenient number of interviewees and attempting to justify the selection afterward.

Document the quantitative selection criteria, recruitment procedures, nonparticipation, and final sample composition. During analysis, preserve the relationship between the selected profiles and the qualitative findings.

Most importantly, ensure that the integrated interpretation does not exceed what the sampling design can support. A purposive interview sample can provide substantial explanatory insight without supporting population prevalence estimates or definitive causal conclusions.

Check Your Quantitative-Informed Interview Sampling Plan

Frequently Asked Questions About Selecting Interview Participants From Quantitative Results

Select Participants Because They Can Illuminate the Quantitative Findings

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.