03 · What You Need to Know
The Defining Quality Problem in Mixed Methods Is Integration
Mixed-methods research combines qualitative and quantitative approaches within an intentional design so that their relationship contributes to answering the research problem.
APA's Mixed Methods Article Reporting Standards require researchers to identify the mixed-methods design and describe the integration of qualitative and quantitative data. The standards treat mixed methods as more than the presence of two datasets.
Similarly, the Mixed Methods Appraisal Tool evaluates whether there is an adequate rationale for using mixed methods, whether qualitative and quantitative components are effectively integrated, whether the outputs of integration are interpreted adequately, whether divergences are addressed, and whether each component meets the quality criteria of its own methodological tradition.
First Ask Why You Need Mixed Methods at All
Mixed methods should solve a research problem that one approach alone would address less completely.
A researcher might need quantitative evidence about how common a phenomenon is and qualitative evidence about how participants experience it. One component might explain an unexpected result from another. Qualitative findings might help develop an instrument that is later tested quantitatively. Researchers might compare different forms of evidence to identify convergence, contradiction, or contextual variation.
The precise rationale matters.
“To obtain a more comprehensive understanding” is often true but too vague by itself. What is incomplete about one method? What specific contribution does the other component make? How will combining them change what can be concluded?
The MMAT treats an adequate rationale for using a mixed-methods design as a quality criterion in its own right.
Each Component Must Be Rigorous on Its Own Terms
Mixed methods does not relax the standards of either methodological tradition.
If the quantitative component uses a biased sample, poorly supported measures, inappropriate statistical analysis, or an inadequate comparison, qualitative findings cannot automatically repair those problems.
Likewise, a methodologically superficial qualitative component does not become rigorous because it accompanies an excellent experiment or survey.
The MMAT explicitly asks whether the different components of a mixed-methods study adhere to the quality criteria of their respective methodological traditions.
Component quality
Are the quantitative and qualitative parts methodologically defensible according to the standards relevant to each approach?
Integration quality
Are the components combined in a way that meaningfully answers the mixed-methods research question and supports the overall interpretation?
A study needs both.
The Quantitative Component Still Needs Quantitative Rigor
The quantitative part should be evaluated according to the design being used.
Relevant concerns may include sampling, measurement, reliability and validity evidence, confounding, randomization where applicable, missing data, statistical assumptions, model specification, precision, and whether conclusions stay within the design's inferential limits.
Nothing about calling the overall project mixed methods changes those requirements.
If a cross-sectional survey cannot support a causal conclusion on its own, adding interviews does not automatically transform the association into a causal effect.
The Qualitative Component Still Needs Qualitative Rigor
The qualitative part likewise requires a coherent methodological approach, appropriate sampling or data-source selection, sufficient engagement with the phenomenon, transparent analysis, reflexivity where relevant, and interpretations grounded in the research material.
Depending on the tradition, researchers may frame these concerns through credibility, dependability, confirmability, transferability, methodological integrity, or other criteria.
The relevant principles of rigor in qualitative research remain important inside a mixed-methods design rather than disappearing once quantitative data are introduced.
Integration Is What Makes the Study Mixed Methods
If the qualitative and quantitative components never meaningfully interact, the project may amount to two parallel studies sharing a topic.
Integration can occur at several points in a mixed-methods design. One component may inform sampling for another. Qualitative findings may help construct a survey instrument. Quantitative results may determine which participants are selected for follow-up interviews. Findings may be compared or merged during analysis. Integrated displays may place qualitative and quantitative results together so that their relationships become visible.
Recent work updating the GRAMMS reporting guidance reiterates that integration is considered a defining feature of mixed-methods research and can occur at sampling, data-collection, analytic, and other levels.
Integration Should Answer a Question, Not Merely Occur Somewhere
Researchers sometimes demonstrate integration by stating that “the quantitative and qualitative findings were integrated during interpretation.”
That description is too thin to evaluate.
What was compared? How were the datasets connected? Did one explain the other? Were qualitative categories compared with quantitative patterns? Did one component identify cases for deeper analysis? What new conclusion emerged because the evidence was combined?
Integration becomes methodologically meaningful when its contribution to the research question is visible.
Timing and Priority Should Follow the Research Purpose
Mixed-methods designs vary in sequence and emphasis.
In a convergent design, qualitative and quantitative components may be conducted during a similar period and then brought together. In an explanatory sequential design, quantitative results may come first and qualitative inquiry may then help explain particular patterns. In an exploratory sequential design, qualitative work may help develop concepts, measures, or hypotheses that are subsequently examined quantitatively.
One component may receive greater emphasis, or the components may be weighted relatively equally.
No sequence is inherently more rigorous. The important question is whether timing, priority, and integration correspond to the mixed-methods purpose.
A Joint Display Can Make Integration Visible
Joint displays organize qualitative and quantitative evidence together so that researchers and readers can examine relationships between them.
For example, a table might present quantitative results for different participant groups alongside qualitative explanations of why those groups experienced an intervention differently.
A joint display is not merely an attractive presentation technique. When designed analytically, it can help researchers identify convergence, divergence, expansion, and new integrated interpretations.
However, inserting numerical results and quotations into adjacent columns does not automatically create integration. The display should support an analytic comparison or inference.
Agreement Between Components Is Not Required
Researchers sometimes assume that mixed methods is strongest when qualitative and quantitative findings tell the same story.
Convergence can certainly strengthen an interpretation. But disagreement may be equally informative.
Suppose a survey shows high faculty confidence in using AI tools, while interviews reveal repeated uncertainty about evaluating AI-generated information. The two findings may appear inconsistent until researchers recognize that self-rated operational confidence and critical evaluative competence are different dimensions.
The discrepancy can refine the interpretation rather than undermine the study.
The MMAT specifically includes adequate handling of divergences and inconsistencies between qualitative and quantitative results among its criteria for mixed-methods studies.
Watch Out
Do not force qualitative and quantitative findings into agreement merely to produce a tidy conclusion. A credible mixed-methods interpretation should explain meaningful divergence rather than conceal it.
The Final Interpretation Should Be More Than Two Separate Conclusions
A quantitative component might conclude that an intervention improved performance. A qualitative component might conclude that participants found the intervention difficult to use.
Reporting those conclusions sequentially is useful, but integration asks what they mean together.
Perhaps the intervention produces measurable benefit despite poor usability. Perhaps the effect is concentrated among participants who overcome particular implementation barriers. Perhaps improved performance comes with additional workload that affects sustainability.
These integrated conclusions are sometimes described as meta-inferences: interpretations produced by considering the components together.
The MMAT explicitly evaluates whether the outputs of integration are adequately interpreted, reinforcing that rigorous mixed methods requires reasoning across components rather than merely presenting each one correctly.
One Component Cannot Automatically Validate the Other
Qualitative and quantitative findings can complement or challenge each other, but researchers should avoid saying that one “validates” the other without specifying what that means.
Interviews cannot repair measurement bias in a survey simply because participants discuss similar topics. A statistically significant association cannot establish that an interpretive qualitative theme is objectively true. Agreement between methods can strengthen particular interpretations, but only when the evidence produced by each method is itself appropriate for that inference.
The broader principle from valid and defensible research design still applies: evidence should support the claim being made rather than acquiring authority simply because several methods were used.
Contradictions Should Be Investigated, Not Averaged Away
When components disagree, several possibilities deserve consideration.
The methods may be examining different aspects of the phenomenon. The samples may differ. Timing may matter. One measure may operationalize the construct differently from the qualitative inquiry. The discrepancy may expose subgroup heterogeneity, contextual differences, or weaknesses in one component.
Mixed methods provides an opportunity to investigate these possibilities.
Simply declaring that “overall, the findings were generally consistent” can waste one of the most analytically valuable features of the design.
Mixed-Methods Sampling Requires Its Own Logic
The quantitative and qualitative components may involve the same participants, overlapping samples, nested samples, or entirely different groups.
Those relationships should be justified.
In an explanatory sequential study, researchers might deliberately select interview participants based on quantitative results, perhaps choosing cases with unusually high, low, or unexpected outcomes. In another study, independent samples may be appropriate because the components address different levels of the research problem.
The question is how the sampling relationship supports integration. If the two samples differ substantially, researchers should consider what those differences mean when comparing findings.
Mixed Methods Does Not Mean “More Data Must Be Better”
Combining methods increases methodological demands.
Researchers need expertise in both traditions, enough resources to conduct each component properly, and sufficient analytic capacity to integrate them. A weak survey plus a few interviews is not necessarily stronger than one well-designed study using the method best suited to the question.
Mixed methods should therefore be chosen because integration is necessary for the research problem, not because two methods look more comprehensive in a proposal.
Reporting Standards Make the Integration Auditable
APA provides specific Mixed Methods Article Reporting Standards alongside its quantitative and qualitative standards. These require authors to identify the mixed-methods design and report information needed to understand the relationship between components.
GRAMMS has also been used as a reporting framework for mixed-methods studies. As of 2026, an international group is developing GRAMMS 2.0 because mixed-methods methodology has evolved considerably since the original guidance and persistent weaknesses in integration remain a concern.
These are reporting resources, not substitutes for methodological quality. A beautifully reported incoherent mixed-methods design remains incoherent, although at least reviewers will be able to diagnose the problem without methodological archaeology.
