What Happens When Quantitative and Qualitative Findings Conflict in Mixed-Methods Research?

What Should You Do When Your Numbers and Interviews Tell Different Stories?

Your survey indicates that most students are satisfied with an online learning platform. Yet during interviews, several students describe frustration, limited engagement, and difficulties completing learning activities. Which finding should you believe?

Perhaps your quantitative analysis suggests that an educational intervention improves performance, while participants describe experiences that seem inconsistent with that conclusion. Or your statistical model identifies a positive relationship between two variables, but qualitative accounts suggest that the relationship is more complicated.

Such differences can be unsettling, particularly when a mixed-methods study was justified partly on the expectation that combining evidence would produce a clearer understanding.

However, integration does not require agreement. The methodological task is to determine whether the findings genuinely contradict one another, investigate why they differ, and establish what conclusions the combined evidence can reasonably support.

Conflicting Findings Should Be Investigated, Not Automatically Reconciled

Understanding and Investigating Divergent Mixed-Methods Findings

What Counts as a Conflict Between Quantitative and Qualitative Findings?

A conflict occurs when findings from different methodological components appear to support incompatible interpretations of a phenomenon.

For example, quantitative evidence may suggest that an intervention is associated with improved learning outcomes, while qualitative participants describe the intervention as ineffective or unhelpful.

However, these findings are not necessarily contradictory. Measured performance and perceived usefulness are different constructs. An intervention could improve performance while remaining difficult or unpleasant to use.

Researchers must therefore distinguish apparent disagreement from genuine incompatibility.

A genuine contradiction is more plausible when both components address sufficiently comparable claims, populations, time periods, and conditions but produce incompatible evidence.

Even then, the researcher should examine whether the findings have been interpreted at an appropriate level of specificity.

Convergence, Complementarity, and Divergence Are Different Relationships

Mixed-methods integration can reveal several relationships between quantitative and qualitative findings.

These relationships are useful analytical descriptions rather than mutually exclusive classifications. Different findings within the same study may converge in some respects and diverge in others.

Greene, Caracelli, and Graham (1989) identified initiation as one purpose of combining methods, recognizing that contradictions and unexpected findings can stimulate reconsideration of existing interpretations.

Consequently, divergence may represent a substantive contribution rather than merely an obstacle to be eliminated.

Are the Components Actually Investigating the Same Construct?

One of the first questions to examine is whether the quantitative and qualitative components concern the same underlying concept.

Suppose a questionnaire asks students to rate their overall satisfaction with an online course, while interviews investigate difficulties with specific learning activities.

High overall satisfaction and frustration with particular activities can coexist. The findings concern different levels of evaluation.

Similarly, platform login frequency is not equivalent to meaningful learning engagement. Quantitative records of frequent access may coexist with qualitative descriptions of passive participation.

Before declaring a contradiction, researchers should examine the operational definitions, questionnaire items, interview questions, and analytical categories used in each component.

A difference in what was measured or explored may explain the apparent disagreement without requiring either finding to be rejected.

Could Sampling Differences Explain the Disagreement?

Quantitative and qualitative components often use different sampling strategies and may involve different participants.

A quantitative survey might include a broad sample of students, while qualitative interviews deliberately select participants who experienced difficulties.

If the interview sample overrepresents dissatisfied students by design, their accounts should not be interpreted as evidence that the survey's overall satisfaction estimate is necessarily incorrect.

The interviews may instead reveal important experiences that are relatively uncommon or insufficiently captured by the questionnaire.

Researchers should examine how participants were selected, which subgroups are represented, and whether the evidence supports population-level or case-level conclusions.

Particular caution is needed when comparing individual interview accounts with aggregate quantitative findings. A group average does not describe every member of the group.

Could Timing or Context Account for the Difference?

Findings collected at different times may reflect changes in participants' experiences or the conditions under investigation.

For example, a satisfaction survey administered immediately after an intervention may produce different responses from interviews conducted several months later.

Participants may have encountered new difficulties, changed their expectations, or reconsidered their earlier experiences.

Likewise, changes in institutional policy, course requirements, technology access, or intervention implementation may affect the comparability of evidence collected across phases.

Researchers should establish whether the components describe the same period and conditions before treating their findings as directly contradictory.

Could Quantitative Measurement or Analysis Be Responsible?

Quantitative findings may be affected by measurement limitations, sampling error, missing data, model specification, or inappropriate statistical assumptions.

For example, a satisfaction questionnaire may contain items that do not adequately represent the dimensions students discuss during interviews.

Alternatively, an overall association may conceal substantial variation across relevant subgroups.

Researchers should examine the instrument's intended interpretation, relevant measurement evidence, descriptive distributions, uncertainty estimates, and analytical decisions.

Where justified, sensitivity analyses or subgroup examinations may help determine whether the apparent disagreement reflects an overly broad quantitative interpretation.

However, researchers should avoid conducting unrestricted post hoc analyses merely to find a statistical result that agrees with the qualitative findings.

Could Qualitative Sampling or Interpretation Be Responsible?

Qualitative findings also require critical examination.

Researchers should consider whether the interview sample adequately represents the experiences relevant to the qualitative question, whether interview prompts were leading, and whether the analysis considered accounts that challenged emerging interpretations.

For example, participants may describe dissatisfaction with a platform because interview questions concentrated on difficulties rather than the full range of experiences.

Similarly, a qualitative analysis may overemphasize particularly vivid quotations while giving insufficient attention to contradictory cases.

Reflexivity, transparent analytical procedures, and examination of alternative interpretations can help researchers assess these possibilities.

This does not mean that qualitative research must reproduce quantitative representativeness. Its quality should be evaluated according to the qualitative approach and the claims being made.

How Do You Investigate Disagreement Systematically?

Researchers can use a structured comparison to identify exactly where the findings differ and which explanations are plausible.

This process should not be understood as a requirement to explain away every disagreement. Sometimes the evidence remains genuinely inconsistent, and that uncertainty is itself an important result.

Can Joint Displays Help Examine Conflicting Findings?

Yes. A joint display can organize quantitative results, corresponding qualitative findings, and integrated interpretations in a common analytical structure.

Guetterman, Fetters, and Creswell (2015) describe joint displays as a means of facilitating integrated analysis and communicating insights that emerge from considering different forms of evidence together.

For divergent findings, a joint display can help identify whether disagreement concerns the entire phenomenon or only particular subgroups, constructs, or contexts.

For example, a table might align survey satisfaction ratings with interview themes concerning specific platform features and identify whether the apparent conflict reflects differences in the level of evaluation.

However, the display should not assign agreement or disagreement mechanically. Researchers must interpret what each finding establishes and whether the evidence is genuinely comparable.

Should One Component Be Treated as More Authoritative?

Not automatically. Quantitative and qualitative evidence support different kinds of inference, and neither approach has universal authority over the other.

A well-designed probability survey may provide stronger evidence about population prevalence than interviews with a small purposive sample. Conversely, carefully conducted interviews may provide more relevant evidence about how participants interpret a particular experience than a questionnaire that never examined that issue.

The appropriate judgment depends on the claim being evaluated, the methodological quality of the evidence, and the intended inferential scope.

Even when a study assigns greater priority to one component, findings from the other may reveal important limitations or alternative interpretations.

Priority does not justify automatically dismissing inconvenient evidence.

What If the Findings Remain Genuinely Contradictory?

Researchers should report the contradiction transparently and explain the limits of the available evidence.

For example, if quantitative and qualitative components produce incompatible interpretations of the same construct among comparable participants during the same period, the study may not support a single coherent conclusion.

The integrated interpretation may instead identify competing explanations, methodological uncertainties, or questions requiring additional investigation.

Researchers should distinguish established findings from speculative explanations for the disagreement.

Additional data collection may sometimes be justified, but it should not be undertaken solely to obtain the preferred result.

In sequential studies, researchers may use unexpected findings to guide a subsequent phase. In convergent studies, the disagreement may become an important outcome of the integration itself.

How Does Divergence Affect the Study's Meta-Inference?

A meta-inference is an overarching interpretation developed from the integration of quantitative and qualitative evidence.

When findings converge, the integrated interpretation may identify compatible evidence supporting a particular conclusion.

When findings diverge, the meta-inference may need to explain why the evidence differs, identify the conditions under which each finding applies, or acknowledge unresolved inconsistency.

For example, a study might conclude that an educational platform receives favorable overall satisfaction ratings while particular aspects of its implementation generate substantial difficulties for some students.

This conclusion preserves both forms of evidence without pretending that they address identical questions.

Researchers should avoid using integration to manufacture a more confident conclusion than either component supports.

How Should Conflicting Findings Be Reported?

Reporting should identify the relevant quantitative result, the corresponding qualitative finding, the nature of the disagreement, and the interpretation supported by the combined evidence.

The report should explain whether the conflict reflects differences in measurement, sampling, timing, analytical levels, or substantive interpretation.

Where the explanation remains uncertain, that uncertainty should be stated directly.

Researchers should not omit contradictory interview accounts, selectively report favorable statistical results, or redefine the research question after seeing the findings without disclosure.

GRAMMS reporting recommendations emphasize transparency about mixed-methods design, integration, and the insights obtained from combining approaches.

Clear reporting allows readers to distinguish genuine convergence from interpretations produced through selective presentation.

Investigating Conflicting Satisfaction Findings in Online Learning

The researcher may conclude that overall satisfaction was favorable among surveyed respondents, while qualitative participants identified specific limitations that the overall satisfaction item did not adequately describe.

This interpretation does not require rejecting either component.

However, the researcher should not infer that 82% of students were satisfied with every platform feature, nor that the interview themes describe the prevalence of dissatisfaction across the entire student population.

The apparent contradiction becomes a more precise account of what the different forms of evidence establish.

Common Mistakes When Quantitative and Qualitative Findings Disagree

How to Respond When Your Mixed-Methods Findings Diverge

Begin by identifying the exact claims that appear inconsistent. Avoid describing the entire study as contradictory when the disagreement concerns only one construct, subgroup, or interpretation.

Then examine the methodological conditions under which each finding was produced.

When writing the results and discussion, make the integration visible. A joint display or structured narrative can show the relationship between the findings and the reasoning behind the overarching interpretation.

The objective is not to make the evidence agree. It is to produce the most defensible account of what the evidence establishes, where its interpretations differ, and what remains uncertain.

Check How You Investigated Conflicting Mixed-Methods Findings

Frequently Asked Questions About Conflicting Mixed-Methods Findings

Disagreement Is Something to Investigate, Not Conceal

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.