When Is a Mixed-Methods Approach Actually Necessary?

Do You Actually Need Two Methodological Approaches to Answer One Research Question?

Suppose you want to evaluate an AI-supported formative feedback system used in university writing courses. You plan to compare students' writing performance before and after the intervention. You are also considering interviews because you believe they would make the study more comprehensive.

But what would the interviews establish that the quantitative evidence cannot? Would they investigate students' feedback interpretation, explain an unexpected performance pattern, or identify processes that the outcome measures cannot capture? Or would they simply provide quotations confirming that students liked the system?

These possibilities have different methodological implications.

Mixed-methods research can address questions that require complementary numerical and qualitative evidence. However, collecting two types of data does not automatically create a stronger study. The additional component must contribute something necessary to the intended answer, and the findings must be meaningfully related.

The central decision is whether your research problem requires integration across methodological approaches, rather than whether both approaches are available or academically fashionable.

Mixed Methods Is Justified When Integration Is Necessary for the Intended Answer

How to Determine Whether Mixed Methods Is Methodologically Justified

What Makes a Study Mixed Methods?

Mixed-methods research involves the intentional use and integration of quantitative and qualitative approaches within a study or coordinated research program to address a research problem.

Its defining feature is not simply collecting numerical data and interview transcripts. The researcher must establish how the two components relate and what their combination contributes to the overall inference.

For example, a study might estimate whether an instructional intervention improves academic performance and then investigate why its effects differ across particular groups of students.

The quantitative component provides an estimate or identifies a pattern. The qualitative component investigates experiences, decisions, or processes relevant to interpreting that pattern. Their integration produces a more developed account of the research problem.

By contrast, a survey containing closed-ended and open-ended questions is not automatically a coherent mixed-methods study. Its classification depends on whether the qualitative responses receive a substantive analytical treatment and how they are integrated with the quantitative findings.

What Does Methodological Necessity Mean?

Methodological necessity means that the intended research answer depends on evidence that a single approach cannot adequately provide.

Consider the question: "What proportion of university students use generative AI for academic writing?"

An appropriately designed quantitative survey may answer this question. Interviews could provide additional context, but they are not inherently necessary to estimate the proportion.

Now consider: "How prevalent are different forms of AI-assisted writing, and how do students understand the academic integrity implications of those practices?"

This question combines a population-level estimation task with an interpretive task. Quantitative evidence may establish the distribution of defined practices, while qualitative evidence may investigate how students understand and negotiate integrity-related decisions.

If both contributions are essential to the research purpose, a mixed-methods approach may be justified.

The decision begins by identifying whether numerical evidence, experiential evidence, or both are required.

What Is the Difference Between Using Two Methods and Conducting Mixed-Methods Research?

Researchers sometimes use several data collection techniques without combining different methodological approaches.

For example, interviews, observations, and document analysis may all contribute to one qualitative case study. Using three techniques does not necessarily make the study mixed methods.

Similarly, combining examination scores, platform logs, and structured survey scales may remain entirely quantitative.

Some methodological traditions use the terms multimethod and mixed methods differently. Researchers should define their terminology and make the actual design procedures explicit rather than relying on the label alone.

What Are the Strongest Reasons to Use Mixed Methods?

Several research purposes can justify combining quantitative and qualitative evidence. These rationales may overlap, but they should not be treated as interchangeable.

These rationales are consistent with established discussions of mixed-methods purposes, including the influential work of Greene, Caracelli, and Graham.

However, naming a rationale is not enough. The researcher must demonstrate how the design operationalizes it.

When Is an Explanatory Sequential Design Appropriate?

An explanatory sequential design generally begins with quantitative data collection and analysis, followed by qualitative inquiry intended to explain or elaborate selected quantitative findings.

Suppose a researcher finds that an AI-supported learning intervention improves average examination performance but produces substantially different outcomes across academic programs.

The qualitative phase might investigate how students and instructors experienced the intervention in programs with contrasting results.

Participants or cases may be selected using the quantitative findings, and qualitative questions may be developed to examine particular patterns.

The design requires more than conducting interviews after a survey. The second phase must be meaningfully connected to the first.

Researchers should also avoid assuming that participants' retrospective explanations establish the causal reasons for quantitative differences. The qualitative evidence may support plausible explanations without definitively identifying causal mechanisms.

When the primary purpose is to investigate how qualitative research can explain quantitative findings, the explanatory contribution and its inferential limits should be specified in advance.

In the supplied collection, the corresponding detailed treatment is available under qualitative explanations of quantitative results.

When Is an Exploratory Sequential Design Appropriate?

An exploratory sequential design generally begins with qualitative inquiry, followed by a quantitative component developed from or informed by the initial findings.

For example, researchers may investigate how students understand AI-generated feedback when little is known about the relevant judgment processes.

The qualitative phase could identify concepts and distinctions that inform the development of survey items or assessment tasks.

A subsequent quantitative phase might examine the measurement properties of those items or estimate how selected characteristics are distributed in a larger population.

However, qualitative themes should not automatically be converted into questionnaire items without conceptual scrutiny, cognitive testing where appropriate, and relevant validation procedures.

Likewise, quantitative confirmation of a pattern does not establish that the initial qualitative interpretation was complete or universally applicable.

The justification lies in the developmental relationship between phases, not merely their chronological order.

When Is a Convergent Design Appropriate?

A convergent mixed-methods design typically collects quantitative and qualitative evidence during a broadly similar period, analyzes the components, and brings their findings together for interpretation.

For example, a researcher might measure students' engagement with an AI learning platform while also observing how they interact with its feedback features.

The quantitative data could describe participation patterns, while qualitative observations illuminate how particular forms of engagement occur.

The findings might converge, diverge, or address different dimensions of the phenomenon.

Convergence should not be treated as automatic proof of validity. Two sources can agree because they share a common bias, while disagreement may reveal meaningful differences in what they measure.

The researcher must decide how the findings will be compared and what an integrated interpretation can establish.

What Does Integration Actually Require?

Integration is the deliberate process of relating quantitative and qualitative evidence to produce an interpretation or inference that addresses the overarching research problem.

Fetters, Curry, and Creswell describe integration at different levels, including study design, methods, and interpretation or reporting.

Integration may involve connecting one phase to another through sampling, building an instrument from earlier findings, merging results for comparison, or embedding one component within a broader design.

At the interpretation stage, researchers may use joint displays, integrated narratives, or other analytical structures to examine how the findings relate.

For example, a joint display might place quantitative engagement patterns alongside qualitative accounts of students' learning practices.

The purpose is not decorative presentation. It is to identify what becomes understandable through examining the evidence together.

A study with separate quantitative and qualitative results sections but no substantive connection between their findings may have limited mixed-methods integration.

How Do You Know Whether Integration Adds Genuine Value?

A useful test is to imagine removing one component from the proposed study.

If removing the qualitative component leaves the primary research question fully answered, the qualitative component may be supplementary rather than necessary.

If removing the quantitative component prevents the researcher from estimating the population distribution or intervention effect required by the question, that component is essential.

However, necessity is not always binary. Some research problems can be answered through more than one defensible design, with different levels of explanatory depth.

Mixed methods may be worthwhile when integration materially improves interpretation, instrument development, or practical decision-making, even if a narrower single-method study would also be legitimate.

The researcher should state the additional contribution rather than claiming that mixed methods is inherently superior.

Can Mixed Methods Strengthen Causal Inference?

Mixed methods can contribute to causal investigation by combining effect estimation with evidence about implementation, mechanisms, or context.

For example, a randomized evaluation may estimate an intervention's effect, while qualitative observations investigate how participants encountered and responded to the intervention.

However, adding qualitative evidence does not automatically strengthen the identification of the average causal effect.

If the quantitative component lacks a credible counterfactual, interviews cannot simply repair that weakness.

Likewise, a statistically significant relationship between a proposed mediator and an outcome does not establish the mechanism without appropriate causal assumptions.

Mixed methods can broaden or deepen causal understanding when each component addresses a legitimate inferential task. It should not be presented as a substitute for adequate design within either component.

What Happens When Quantitative and Qualitative Findings Disagree?

Divergence is not necessarily a methodological failure.

Suppose survey results indicate high student satisfaction with an AI-supported advising system, while interviews reveal substantial uncertainty about the reliability of its recommendations.

The findings may concern different aspects of experience. Students can appreciate convenience while remaining skeptical of particular recommendations.

Alternatively, differences may arise from sampling, measurement, timing, or the conditions under which responses were collected.

Researchers should investigate these possibilities rather than forcing the findings into artificial agreement.

A credible integrated interpretation may explain why the evidence differs, identify unresolved tensions, or establish that the initial conceptualization was incomplete.

What Makes Mixed-Methods Research More Demanding?

Mixed-methods studies require sufficient methodological competence and resources to conduct both components appropriately and integrate them coherently.

The quantitative component may require suitable sampling, measurement, and statistical analysis. The qualitative component may require appropriate participant selection, data generation, interpretation, and reflexivity.

Integration adds another layer of planning and analysis.

Researchers may need to coordinate sequential phases, select participants based on earlier results, reconcile different units of analysis, and explain how findings with different inferential purposes relate.

These demands can affect feasibility, timelines, ethical procedures, and the overall quality of the study.

Using mixed methods simply because both types of data are available may create an unnecessarily complex design. Availability is a resource consideration, not a sufficient methodological rationale.

When Is a Single-Method Study More Defensible?

A single-method study may be preferable when one approach can answer the research question adequately and the additional component would not resolve a meaningful evidentiary gap.

For example, a study estimating the prevalence of generative AI use among undergraduate students may be well served by a carefully designed quantitative survey.

A study examining how teachers negotiate professional identity during AI adoption may be better served by a coherent qualitative approach.

Neither study necessarily benefits from adding another methodological component.

Researchers should consider when quantitative research is appropriate and when qualitative research is appropriate before concluding that both are required.

The appropriate choice depends on the evidence necessary to support the intended claim, not the number of methods included.

When Mixed Methods Adds Necessary Evidence to an AI Literacy Study

Suppose students with similar AI literacy scores demonstrate different approaches to evaluating AI-generated information. Some examine supporting sources, while others rely primarily on the apparent fluency of the response.

The qualitative evidence could reveal distinctions that the numerical score does not capture. Conversely, the quantitative component could establish the distribution of assessed performance within the studied population, subject to the sampling and measurement design.

The integrated contribution would concern both measured AI literacy and the practices associated with interpreting AI-generated information.

However, the study would not automatically establish that particular evaluation practices caused higher scores. That causal claim would require additional evidence and assumptions.

Mixed methods is justified here because both components contribute to an overarching question that cannot be answered as fully by either component alone.

Common Misconceptions About Mixed-Methods Necessity

Use Mixed Methods Only When the Integrated Contribution Is Clear

Before choosing mixed methods, identify what each component is expected to establish and what the study will learn by integrating them.

If the integrated conclusion cannot be stated clearly, the design may not yet have a sufficient rationale.

A practical test is to complete this sentence before data collection: "The quantitative findings will establish..., the qualitative findings will establish..., and integrating them will allow us to understand..."

If the final part merely repeats the first two, the rationale for integration may need further development.

Mixed methods should be selected because the research problem demands the combined evidence, not because methodological variety is treated as a marker of quality.

Before Choosing Mixed Methods, Check These Requirements

Frequently Asked Questions About When Mixed Methods Is Necessary

Mixed Methods Is Justified by the Need for Integration, Not Methodological Variety

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.