When Is Mixed Methods Actually Mixed Methods, and When Are You Just Collecting Two Types of Data?

If Your Study Has Numbers and Interviews, Is It Automatically Mixed Methods?

You administer a questionnaire containing Likert-scale items and one open-ended question. Is that mixed methods? You conduct a survey and then interview several participants. Is that enough? What if you analyze numerical results and interview transcripts but report them in completely separate sections?

The presence of both quantitative and qualitative data is necessary for many definitions of mixed methods research, but it is not sufficient to make the study methodologically mixed.

The more consequential question is what happens between the components. Mixed methods research deliberately brings qualitative and quantitative approaches into relation so that their combination contributes to answering the research problem. Without meaningful integration, you may simply have two types of data occupying the same study.

Mixed Methods Requires Integration, Not Just Coexistence

The Defining Question Is What the Qualitative and Quantitative Components Do Together

Mixed Methods Is More Than “Quantitative + Qualitative”

Mixed methods research intentionally combines elements of quantitative and qualitative approaches within an integrated program of inquiry. Definitions vary across the methodological literature, but integration or mixing is widely treated as a central distinguishing feature.

This distinction matters because two independently conducted components do not automatically produce mixed methods simply because they appear under the same project title. If the quantitative component answers one question, the qualitative component answers an unrelated question, and the findings never inform or interact with one another, calling the study mixed methods may overstate the methodological integration.

A useful diagnostic question is: What can I understand from combining these components that I could not understand as well by keeping them separate?

If there is no convincing answer, the study's mixed methods rationale probably needs further development.

Integration Is the Methodological Work That Connects the Components

Integration refers broadly to deliberately relating qualitative and quantitative components so they contribute to a combined understanding of the research problem. Methodologists have described integration as occurring in different dimensions and at different stages rather than as one single technique.

For example, one component may inform whom you sample for another. Quantitative findings may identify a surprising pattern that qualitative interviews subsequently investigate. Qualitative findings may help develop a quantitative instrument. Separate datasets may be brought together during analysis. Findings may be compared and interpreted jointly to identify convergence, divergence, expansion, or new explanations.

The important feature is intentional connection. Integration should follow the study's purpose rather than appear as a paragraph added near the end stating that “the quantitative and qualitative findings support each other.”

Collecting an Open-Ended Survey Response Does Not Automatically Create Mixed Methods

Suppose a questionnaire contains 30 closed-ended items and concludes with, “Please provide any additional comments.” Researchers calculate descriptive statistics for the closed items and quote several comments in the discussion.

There are quantitative and textual data, but that alone does not demonstrate a rigorous qualitative component or meaningful mixed methods integration.

The same caution applies when researchers automatically classify any survey containing both closed and open questions as mixed methods. The qualitative component should have a defensible purpose, appropriate data-generation strategy, systematic analysis, and meaningful relationship to the quantitative component.

A Survey Followed by Interviews Can Be Mixed Methods, but the Connection Must Matter

Imagine surveying 1,000 university students about generative AI use and academic engagement, then interviewing 20 students.

That sequence could constitute mixed methods if the interviews are intentionally connected to the quantitative phase. Perhaps participants are selected because they represent unexpected quantitative patterns. Perhaps the interviews investigate why a statistical relationship appeared, why an expected relationship did not appear, or how students interpret a pattern that numerical measures alone cannot explain.

Now imagine that the survey measures AI use and engagement while the interviews ask unrelated questions about students' general campus experiences. Both forms of data exist, but their coexistence does not create a coherent mixed methods inquiry.

Connection is therefore substantive, not chronological. Doing one method after another is not enough.

Integration Can Occur at Different Points

A study does not necessarily integrate at every possible point. The appropriate integration strategy follows the mixed methods design and purpose.

Integration Can Involve Connecting, Building, Merging, and Embedding

Mixed methods literature uses several terms to describe how components interact. Although terminology varies somewhat among authors, several strategies are particularly useful for researchers to recognize.

Connecting occurs when one component influences sampling for another. For example, quantitative results may identify participants for qualitative follow-up.

Building occurs when findings from one component inform the data collection of another. Qualitative interviews might help researchers construct questionnaire items, or survey findings might identify issues requiring deeper interview questions.

Merging brings qualitative and quantitative results together for comparison or combined analysis. Researchers may examine whether the findings converge, contradict, or illuminate different aspects of the same phenomenon.

Embedding places one form of evidence within a larger design serving another primary purpose. For example, qualitative data may be embedded within an intervention study to investigate implementation or participant experiences.

These strategies are methodological mechanisms, not boxes to tick. Researchers should identify the integration strategy because it solves a particular research problem.

Mixed Methods Should Have a Mixed Methods Question or Integrative Purpose

A strong mixed methods study makes the reason for integration visible in its questions or objectives. The study may contain quantitative questions and qualitative questions, but it should also have an overarching mixed methods purpose that explains how the components contribute jointly.

For example:

Quantitative question: What is the relationship between frequency of generative AI use and academic self-efficacy among university students?

Qualitative question: How do students describe ways in which generative AI affects their confidence when completing academic tasks?

Integrative question: How do students' accounts help explain or contextualize the quantitative relationship between generative AI use and academic self-efficacy?

The final question makes the relationship between components explicit. Without that integrative purpose, researchers may conduct two respectable studies that happen to share participants but never actually become one mixed methods inquiry.

Timing and Integration Are Different Decisions

Qualitative and quantitative components may occur sequentially or concurrently. In a sequential design, one component precedes and usually informs the next. In a concurrent or convergent arrangement, components may be conducted during a similar phase and brought together subsequently.

Neither sequence is inherently more mixed. What matters is why the timing was chosen and how the components interact. The decision about whether qualitative and quantitative components should occur sequentially or concurrently should therefore follow the intended function of integration.

Equal Importance Is Not Required

Mixed methods does not require qualitative and quantitative components to receive exactly equal sample sizes, pages, resources, or analytical emphasis. Some designs prioritize one component and use the other in a supplementary but methodologically meaningful role.

The important issue is whether each component is rigorous enough for its intended function and whether the relationship between them is justified.

A small qualitative component is not automatically tokenistic, just as a large one is not automatically meaningful. If ten carefully selected interviews explain a consequential quantitative pattern, their contribution may be substantial. If fifty interviews are collected but never integrated with the quantitative findings, their larger number does not solve the design problem.

Disagreement Between Findings Is Not a Failure of Mixed Methods

Researchers sometimes expect qualitative and quantitative findings to confirm one another. When they do not, the temptation is to privilege one dataset or explain the discrepancy away.

Yet divergence can be one of the most informative products of integration. A survey might show high reported satisfaction while interviews reveal substantial frustration under particular circumstances. That apparent contradiction may indicate differences in measurement, subgroups, context, interpretation, or dimensions of the phenomenon that the survey did not capture.

Integration therefore involves examining agreement and disagreement rather than treating qualitative evidence as decorative confirmation of quantitative results.

Mixed Methods Is Not Automatically Better Than a Single-Method Study

Using two methodological approaches can provide breadth, depth, explanation, complementarity, or other forms of insight. It also increases demands on design, expertise, sampling, data management, analysis, integration, reporting, and often time.

If the research question can be answered adequately with one approach, adding another may create work without improving the inference. Methodological pluralism is useful when the problem needs it, not as an ornament on the methods section.

This is another instance in which a more informative but more complex design should justify the additional burden it creates.

Two Types of Data Can Either Remain Separate or Become Mixed Methods

Now remove the connection and integration. Suppose the same researcher conducts the survey, separately interviews whichever faculty members happen to volunteer, analyzes the two datasets independently, and reports two unrelated sets of findings.

The project still contains quantitative and qualitative data. What has disappeared is the methodological relationship that would make their combination analytically consequential.

Common Mistakes When Calling Research Mixed Methods

Design the Integration Before You Collect the Two Types of Data

If you are planning mixed methods research, do not begin with “I will use a survey and interviews.” Begin with the gap that one methodological approach cannot adequately address.

Then identify the point or points of integration before data collection begins. Will one component determine sampling for another? Will findings build an instrument or interview protocol? Will datasets be merged during analysis? Will you use a joint display to examine relationships across findings? What integrated conclusion will the design make possible?

Those questions also clarify timing. Once you know what one component needs from the other, it becomes easier to decide whether one component must logically come first or whether they can proceed concurrently.

Finally, plan enough expertise and space for both components. Mixed methods manuscripts can become strangely asymmetrical when one component receives ten pages of rigorous analysis and the other receives a paragraph beginning “participants were also interviewed.” Integration cannot rescue a component that was never adequately designed.

Before Calling Your Study Mixed Methods

Questions Researchers Ask About Mixed Methods Research

The “Mixed” Part Is What Happens Between the Methods

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.