Did One Method Explain, Expand, or Contradict the Other?

What happened when the two forms of evidence met?

After reading the quantitative and qualitative results of a mixed methods study, you should be able to ask more than whether they “agree.” Agreement is only one possible relationship between two forms of evidence.

The qualitative component might explain an unexpected statistical pattern. It might expand a numerical finding by revealing variation hidden inside an average. Quantitative evidence might indicate how broadly a qualitatively identified pattern appears within the studied sample. Or the two components might contradict each other in a way that forces researchers to reconsider their original interpretation.

The important question is therefore not simply whether both methods produced findings. It is what researchers learned by examining the relationship between those findings.

Look for the function of the relationship between the findings

Mixed methods findings can relate in several meaningful ways

Do not reduce integration to agreement versus disagreement

It is tempting to treat mixed methods integration as a simple validation exercise: if the numbers and interviews agree, confidence increases; if they disagree, something has gone wrong.

That view is too narrow.

Fetters, Curry, and Creswell describe the “fit” of integration in terms of how qualitative and quantitative findings cohere. Their framework recognizes confirmation, expansion, and discordance as different relationships that can emerge when findings are brought together.

This broader view matters because different mixed methods designs are created for different purposes. An explanatory sequential design may deliberately use qualitative inquiry to explain selected quantitative results. An exploratory sequential design may use qualitative findings to build a later quantitative component. A convergent design may compare two forms of evidence collected during a similar period. The expected relationship should therefore follow the study's design and purpose.

Explanation asks why a quantitative pattern may have occurred

Suppose a statistical analysis shows that students using a learning platform more frequently have higher average engagement scores, but the association is much weaker for one subgroup. Researchers then interview members of that subgroup to investigate what differs about their experience.

The interviews might reveal that frequent platform use includes very different activities. Some students engage deeply with learning materials, while others repeatedly log in because they are struggling to locate required resources.

The qualitative evidence may therefore provide a plausible explanation for why a seemingly straightforward quantitative indicator behaves differently across participants.

That explanation should still be interpreted according to what the qualitative evidence can support. Interviews may illuminate mechanisms or participant perspectives, but they do not automatically establish that the proposed explanation caused the quantitative association.

Expansion adds dimensions that the other method did not capture

Sometimes the two methods are not intended to answer the same immediate question. Instead, one expands the other.

Imagine an intervention study showing a modest improvement in average test scores. Interviews then reveal that participants experienced the intervention differently depending on access to technology, instructor support, or how they incorporated the intervention into their study routines.

The qualitative findings do not merely repeat the quantitative outcome in words. They broaden the interpretation by revealing variation, conditions, processes, or meanings not represented by the average effect.

Mixed methods scholarship treats expansion as one possible relationship between integrated findings. This can be especially useful when a single numerical summary would otherwise conceal meaningful heterogeneity.

Complementarity means the findings answer different parts of a larger problem

Two findings may fit together without one directly explaining the other.

For example, quantitative evidence may estimate how frequently teachers use generative AI for different tasks, while interviews examine how teachers decide whether a particular use is pedagogically or ethically acceptable. One component describes behavioral patterns; the other examines reasoning surrounding those behaviors.

Their relationship is complementary because each contributes something necessary to the larger research problem.

This works only if each component addresses a meaningful part of the research question. Two unrelated findings do not become complementary merely because they concern the same broad topic.

Qualification can change how broadly a finding should be interpreted

A qualitative component may reveal that an apparently general quantitative finding has important boundaries.

Suppose 82% of respondents report that a new academic advising system is useful. Interviews may show that “useful” means very different things to different students. Some value faster administrative processing, while others value personalized advice. A smaller group may report high usefulness scores despite describing serious accessibility problems.

The qualitative findings do not necessarily overturn the 82% result. They change what that percentage can reasonably be taken to mean.

That is a substantive contribution. Integration sometimes makes a conclusion more conditional rather than more decisive.

Contradiction can expose something the separate analyses missed

Suppose survey respondents report high confidence in using an educational technology. During interviews, however, many of those same participants describe substantial uncertainty when asked to explain how they evaluate the accuracy of its outputs.

The apparent contradiction may have several explanations. The survey item and interview questions may operationalize confidence differently. Participants may interpret the response scale differently. Social desirability may influence one form of response. Confidence may be domain-specific rather than general. Or the discrepancy may reveal a genuinely important distinction between perceived competence and demonstrated reasoning.

The contradiction is therefore not merely an inconvenience to be averaged away. It becomes something to investigate.

When disagreement is substantial, the next methodological question is what researchers should do when quantitative and qualitative findings disagree.

The relationship should be demonstrated rather than declared

Authors sometimes write that qualitative findings “supported” the quantitative results without showing precisely how.

Look for evidence. Which quantitative result corresponds to which qualitative theme? What feature of the qualitative evidence explains or qualifies the numerical pattern? Where do cases diverge? What integrated inference follows from the comparison?

Joint displays can be particularly useful because they place findings into explicit relationships. But a joint display is only as useful as the reasoning it contains. Two columns labeled “quantitative” and “qualitative” do not automatically create integration.

The larger test remains whether the study actually integrates the two components rather than merely placing their findings near each other.

Sample relationships affect how confidently findings can be related

Interpretation also depends on who contributed the evidence.

If qualitative interviews were deliberately selected from quantitative profiles, researchers can examine those profiles at the case level. If the components use different samples, the relationship may instead operate at the population, setting, organizational, or conceptual level.

Neither arrangement is inherently inferior, but researchers should not imply person-level explanation when the sampling design cannot provide it. This is why evaluating whether the quantitative and qualitative samples were appropriately connected is important before accepting claims that one component explains another.

Integration should affect the final interpretation

The relationship between findings matters only if researchers allow it to change their reasoning.

If qualitative evidence reveals exceptions to a quantitative pattern, the final conclusion should preserve those exceptions. If findings contradict one another, the conclusion should not quietly report only the more convenient result. If qualitative evidence provides a plausible explanation rather than causal proof, the language should remain appropriately cautious.

A useful test is to compare the conclusion you would draw from the quantitative component alone with the conclusion you would draw after reading both components. If nothing changes, the integrated contribution may be modest.

The same findings can converge in one respect and diverge in another

Common mistakes when interpreting relationships between mixed methods findings

Ask what changed when the findings were considered together

When reading a mixed methods study, reconstruct the relationship between specific findings rather than accepting broad claims that the components “supported each other.”

The best integrated interpretation is not necessarily the neatest one. Sometimes the contribution of mixed methods research is precisely that a simple conclusion becomes less simple after another form of evidence is considered. Methodological inconvenience has occasionally been known to improve a paper.

The next question is whether that integration can legitimately support a conclusion that neither component supports on its own, which requires careful attention to how the combined inference is constructed.

Check what the findings actually do to one another

Questions about relationships between mixed methods findings

Good integration asks what each finding does to the interpretation of the other

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.