Mixed Methods Research

Understand what mixed methods brings together

Mixed methods research purposefully combines qualitative and quantitative strands and integrates them within a coherent study or connected programme of inquiry. Integration makes their relationship contribute to answering the research question.

Suppose a study finds that a checking prompt reduces sharing in a simulated messaging task. That result estimates a difference under the study conditions. Interviews might then explore what participants noticed, how they interpreted the prompt, and why some still shared. A combined interpretation can connect the measured outcome with accounts of the decision process and identify questions for a later study.

Look for the point of integration

A study might use survey findings to select interview participants, use fieldwork to build a questionnaire, or compare independently developed findings through a joint display. The plan should specify what information crosses between strands and what the researcher does with it.

“Multimethod” has several uses in methodological writing. A project using interviews and observation may combine multiple qualitative methods. In this chapter, “mixed methods” refers specifically to the purposeful integration of qualitative and quantitative inquiry. Describe the actual design and procedures alongside its label.

Explain why the question needs both strands

A mixed methods rationale identifies the understanding that the combination is intended to produce. Begin with the problem, then specify what each strand contributes and how bringing them together helps.

The lecture introduces five purposes associated with the influential framework of Greene, Caracelli, and Graham. Several can operate within one study.

On a narrow screen, scroll the table sideways.

Write the added value in a sentence

For the checking-prompt study, an experiment can estimate a task-specific assignment effect. Interviews can explore participants’ interpretations and decisions. Linking the two can help generate a more precise account of how the intervention is experienced and what should be investigated next.

Judge whether the extra work serves the inquiry

Mixed methods can develop measures, investigate puzzling results, and connect outcomes with context. It also demands time, suitable expertise, and a workable plan for integration. Assess whether you can conduct both strands well and whether their connection answers a consequential question. A focused study may need one well-chosen approach.

Align questions, theory, and perspective

A coherent mixed methods study gives each strand a clear task and states a question about their relationship. The questions should address the same substantive problem at compatible levels.

A quantitative hypothesis can specify an expected contrast, such as a lower sharing rate under the prompt. Qualitative questions can remain open to interpretations the team has not anticipated. The mixed methods question identifies the intended analytical connection.

Let theory guide an explicit set of expectations

A conceptual framework might connect message cues, perceived credibility, group expectations, and sharing decisions. Use it to explain what is measured, what the interview guide explores, and what relationships the combined analysis will examine. Qualitative findings may also reveal a missing concept or challenge the proposed relationship.

The role and timing of theory vary with the design. An exploratory study can develop concepts before testing a measure. A different study can begin with a developed framework and examine it through complementary evidence. Explain how the strands’ concepts correspond and where their perspectives differ.

State the perspective behind the choices

Pragmatism often guides mixed methods through the problem, the consequences of inquiry, and the usefulness of the resulting knowledge. It still requires a reasoned account of methods and inference. A dialectical stance treats differences in assumptions and interpretations as resources for investigation. A transformative or participatory orientation can place collaboration, whose knowledge counts, and the intended use of findings at the centre of the design.

Researchers’ positions also shape the inquiry. A lecturer investigating students’ checking behaviour may expect the prompt to educate, while students may experience it as an instruction to give the approved response. Record how that relationship enters recruitment, measurement, interviewing, and interpretation.

Distinguish the three core designs

A mixed methods design organises the strands’ timing, priority, and connections. The three core designs provide starting structures that can be adapted to the research problem.

Convergent: develop both strands, then compare

In a convergent design, qualitative and quantitative work proceeds in a similar period. Develop each analysis appropriately, then bring findings together around shared questions. For example, a survey about newsroom verification practices can be compared with interviews or observations concerning those practices in the same period.

Explanatory sequential: follow a numerical finding into its context

Begin with quantitative data and analysis. Use selected results to guide qualitative follow-up, then integrate the accounts with the initial findings. Follow-up can investigate an important contrast, an unexpected result, or a result with substantial uncertainty; it need not wait for statistical significance.

Exploratory sequential: develop concepts, then investigate their wider pattern

Begin with qualitative inquiry into experiences, practices, or local meanings. Use the findings to develop a measure, intervention, or set of propositions for subsequent quantitative investigation. Return to the qualitative concepts when interpreting the later results.

Specify priority and any larger framework

Priority concerns the contribution of each strand to the overall purpose; it is not determined by sample size or which phase comes first. In a diagram, → can indicate sequence and + parallel timing. If capitalisation indicates priority, define that convention in a legend.

Core designs can operate within an intervention study, an intensive case study, a participatory project, or a multiphase programme. An embedded component addresses a specified role within a larger design. Explain its integration points and contribution. The NIH guide and Fetters, Curry, and Creswell develop these possibilities.

Plan how integration will happen

Integration becomes practical when you name the information that moves between strands and the analytical decision it changes.

A study can combine several of these operations. The worked example connects an experiment with an interview sample, builds the interview guide from task results, and merges findings during interpretation.

Choose a unit and a place for the comparison

Integration may occur around a participant, a newsroom, a theme, an intervention phase, or a theoretical proposition. Decide what makes the evidence belong together. A university-level interview about policy and a student-level sharing outcome may provide complementary context, with different units and scopes.

Use transformation when it serves a clear purpose

Sometimes researchers turn qualitative codes into numerical indicators or use quantitative profiles to structure a qualitative case comparison. Document the transformation and retain access to the original material. Define what is counted: participants, passages, events, or cases.

For example, “a participant discussed time pressure at least once” differs from “the number of time-pressure references in a transcript”. Neither automatically estimates how common that concern is in a wider population. Describe how the transformation changes the information and what interpretive detail it leaves behind.

Justify both samples and their relationship

Each strand needs a sampling rationale suited to its question. The mixed methods plan also explains how the selected people, cases, or materials relate across strands.

A quantitative population estimate requires a credible population definition, selection strategy, and treatment of participation. An experiment requires suitable assignment, outcomes, and a sample-size justification for its intended contrast. Qualitative selection seeks relevant experiences, depth, and useful variation. Explain the evidence each sample can provide.

On a narrow screen, scroll the table sideways.

Plan follow-up while the first phase is being designed

If you need to recontact participants, establish permission and secure linkage from the outset. Define likely selection criteria while allowing the actual first-phase findings to refine the follow-up. Record who was eligible, invited, and interviewed, including refusals where known.

A contrasting interview sample may include participants with different outcomes within each assigned condition. Their accounts can investigate experiences around those outcomes. The selection is an analytical decision, and the interview subset has its own scope.

Give each strand an adequacy rationale. Quantitative size may follow precision or power requirements. Qualitative adequacy depends on the focus, diversity, quality of material, and analytical approach. There is no required numerical ratio between the strands.

Build a workable and reflexive protocol

A mixed methods protocol specifies the procedures within each strand and the work that joins them. It should be feasible for the team, resources, and access available.

• Map the phases: Specify timing, dependencies, decisions, and the point at which each phase can begin.
• Prepare the materials: Develop measures, task instructions, interview guides, observation arrangements, and pilot procedures.
• Plan each analysis: Define quantitative outcomes and contrasts, the qualitative analytical approach, and integration procedures.
• Assign responsibilities: Identify who leads each strand and who will work across them to develop the combined interpretation.
• Reserve integration time: Schedule shared analysis sessions, display building, examination of differences, and writing.
• Document revisions: Distinguish planned steps from changes prompted by emerging findings and explain their reasons.

Work across methodological perspectives

Teams can bring statistical, qualitative, substantive, linguistic, and community knowledge to the same problem. Agree what terms such as “trust”, “effect”, or “checking” mean within each strand. Preserve important differences rather than forcing every account into a common numerical category.

A reflexive record can track how the team’s expectations influence selection and interpretation. If the intervention designers conduct interviews, participants may feel pressure to praise the prompt. Consider interviewer roles and neutral prompts, then examine how the relationship shaped the material.

Plan consent and linkage together

Explain the task, interview, possible recontact, linkage of records, and uses of quotations. Participation in one phase should not silently become agreement to another. Store the linkage key separately where appropriate and restrict access to identifiable material.

Combining sources can make a person more recognisable: a rare task outcome, job role, and distinctive quotation may identify someone even after names are removed. Assess the combined disclosure risk when preparing displays and reports.

A worked study: a prompt before sharing

The worked example follows a numerical comparison into a qualitative investigation of the decision. Its integration plan is established before the experiment and refined through the observed results.

Design: An explanatory sequential mixed methods study within an intervention framework. Four hundred volunteer adult students are individually randomised, 200 to a checking prompt and 200 to a comparison condition, before one simulated sharing task. The study uses a created misleading claim, with appropriate consent, review, and debriefing arrangements. Assume all 400 provide the task outcome.

Phase 1: estimate the defined difference

The primary outcome is whether the participant selects “share” in the task. The planned contrast compares the two assigned groups. The figures below are invented.

An approximate 95% Wald interval for that risk difference is −23.6 to −6.4 percentage points, using independent binomial observations. The estimate concerns assignment to this prompt under these task conditions.

Connect: select experiences for follow-up

With permission for recontact, the team purposively selects contrasting task experiences. Suppose six interviews are completed in each cell below. This is an illustrative plan, with qualitative adequacy evaluated through the focus and material generated.

On a narrow screen, scroll the table sideways.

Phase 2: investigate the accounts

Begin by asking participants to reconstruct the task in their own terms. Probe what they noticed, what seemed credible, what they considered doing, and what happened next. Ask about the prompt without assuming they noticed it or followed it.

Suppose an appropriately conducted qualitative analysis develops two illustrative themes: “A reason to pause” and “Urgency outruns checking.” One fictional participant says, “The reminder made a pause feel sensible.” Another says, “It looked urgent, so I passed it on.” Examine contrasting accounts and the wider interview material before developing these themes.

Build a joint display and an integrated conclusion

A joint display arranges qualitative and quantitative findings around a common analytical question. An interpretation column shows what their relationship contributes.

Guetterman, Fetters, and Creswell describe joint displays as a means of developing integration. Choose rows that suit the question: cases, concepts, phases, findings, or instrument-development decisions. Link each entry to its source and preserve uncertainty.

On a narrow screen, scroll the display sideways.

Develop a meta-inference

A meta-inference is a conclusion developed through the relationship between strands. It should say what was learned from their connection and how far that claim reaches.

The numerical strand alone leaves the participants’ interpretations unexplored. The qualitative strand alone supplies no estimate of the assigned-group difference. Their integration develops a more specific agenda for explanation and intervention design.

Check the comparison before calling it agreement

A theme about perceived pressure and an estimate of sharing concern related but different things. Ask whether they address the same concept, people, period, and level. Label a relationship as convergence, complementarity, divergence, or an unresolved gap only after examining that fit.

Evaluate the quality of the combined claim

Mixed methods quality rests on the integrity of each strand and the reasoning that joins them. A strong connection makes evidence useful while retaining its limits.

Investigate divergence systematically

Suppose a questionnaire records frequent checking, but interviews describe forwarding without verification. Begin by examining the comparison: the questionnaire may concern general habits over a month, while the interview concerns one urgent event.

1. Check the records. Verify coding, linkage, transcription, scoring, and time periods.
2. Compare the constructs. Ask whether both strands use “checking” to mean the same activity.
3. Compare the samples and situations. Examine selection, nonresponse, setting, and the activity that generated the material.
4. Examine the interpretation. Consider exceptions, plausible alternatives, and what each source can actually establish.
5. Retain or investigate the tension. Revise the claim or identify further evidence needed, documenting the decision.

Use legitimation as a set of practical questions

The lecture uses legitimation for examining the credibility of the mixed methods process and its inferences. Translate this into questions about your design: Do the samples support the intended comparison? Are participants’ meanings and analytical concepts both considered? If data were transformed, is the transformation defensible? Could the sequence have changed responses? Does the combined conclusion exceed what either strand can support?

For the worked experiment, randomisation supports a defined comparison under relevant assumptions. Recruitment limits the population claim; the simulated task limits the setting; purposive follow-up limits the scope of interview patterns. These limits belong within the integrated interpretation.

Plan sensitivity work that addresses a real concern. This could include examining missing outcomes, comparing interpretations across contrasting cases, or checking whether a different defensible coding decision changes the integrated account. Preserve the reasoning in an integration memo.

Appraise the study and report the integration

A clear mixed methods report shows why the strands were combined, how the design connected them, and what the integrated findings add.

Use appraisal and reporting guidance for their purposes

The Mixed Methods Appraisal Tool, version 2018, helps examine methodological quality across qualitative, quantitative, and mixed methods components. Its mixed methods criteria address the rationale, effective integration, interpretation of integrated outputs, treatment of divergence, and quality of the individual components. Report criterion-level judgements with supporting reasons; an overall score can hide which part needs attention.

GRAMMS—Good Reporting of a Mixed Methods Study—offers a useful reporting framework. Its six concerns can be used as prompts:

1. Explain the reason for using mixed methods.
2. Describe the design, including purpose, sequence, and priority.
3. Report the sampling, data generation, and analysis of each component.
4. Explain where and how integration occurred and who participated in it.
5. Discuss limitations created by combining the methods.
6. Describe insights gained through their connection.

The APA JARS–Mixed standards provide further guidance on reporting both strands and their integration. Use the relevant guidance for your discipline and journal. Clear reporting makes the work assessable; the design and evidence still carry the methodological argument.

Organise the paper around the design

An explanatory sequential report can present the quantitative phase, the connection to follow-up, the qualitative phase, and the integrated interpretation. A convergent report can bring findings together by concept. In either case, a design diagram, sampling map, and joint display help readers follow the logic.

In the discussion, distinguish established findings, interpretations, proposed mechanisms, and recommendations for further testing. Make the added contribution visible in the abstract and conclusion as well as the methods.

Learn with the lecture and design workshop

Use the lecture to revisit each design decision. Keep a question sheet, a design sketch, and a blank joint display beside you.

Open the lecture on YouTube. The approximate starting points follow the supplied transcript. Use the chapter’s design definitions when mapping the worked example.

Attempt each task before opening the response

Glossary and further reading

Use these terms to explain what the study combines and how the combination contributes. The readings provide methodological and reporting guidance beyond this introduction.

Strand

A qualitative or quantitative component with its own question, procedures, and analysis.

Integration

The deliberate connection of strands so their relationship contributes to the inquiry.

Convergent design

A design developing strands in a similar period and bringing findings together for comparison and interpretation.

Explanatory sequential

Quantitative findings guide qualitative follow-up, followed by an integrated interpretation.

Exploratory sequential

Qualitative findings shape subsequent quantitative work, with integration across the phases.

Priority

The relative contribution or emphasis of a strand within the overall purpose.

Connecting

Linking strands through selection, such as choosing an interview subset from a survey.

Building

Using findings from one strand to shape a later data-generation approach or intervention.

Merging

Bringing evidence together for comparison and integrated interpretation.

Embedding

Linking a complementary component within a larger research design at specified points.

Joint display

A structured arrangement that links findings and helps develop an integrated reading.

Meta-inference

A conclusion developed through the relationship between qualitative and quantitative evidence.

Legitimation

Examining the quality and credibility of mixed methods procedures and the resulting inferences.

Sources and guided reading

1. Dr. USP. Mixed Methods in 15 Minutes [Video]. YouTube.The accompanying lecture and transcript introduce purposes, designs, integration, reflexivity, quality, and an intervention example.
2. Greene, J. C., Caracelli, V. J., & Graham, W. F. (1989). Toward a conceptual framework for mixed-method evaluation designs. Educational Evaluation and Policy Analysis, 11(3), 255–274. https://doi.org/10.3102/01623737011003255Read for the five purposes and a framework for mixed-method evaluation.
3. Creswell, J. W., Klassen, A. C., Plano Clark, V. L., & Smith, K. C. (2011). Best practices for mixed methods research in the health sciences. National Institutes of Health, Office of Behavioral and Social Sciences Research. Full guide [PDF].A guide to design, rationale, teamwork, resources, and evaluation. This chapter applies general methodological ideas to communication research.
4. Fetters, M. D., Curry, L. A., & Creswell, J. W. (2013). Achieving integration in mixed methods designs—Principles and practices. Health Services Research, 48(6 Pt 2), 2134–2156. https://doi.org/10.1111/1475-6773.12117Develops integration at design, methods, interpretation, and reporting levels.
5. Guetterman, T. C., Fetters, M. D., & Creswell, J. W. (2015). Integrating quantitative and qualitative results in health science mixed methods research through joint displays. Annals of Family Medicine, 13(6), 554–561. https://doi.org/10.1370/afm.1865Use when planning how a display will develop and communicate integrated findings.
6. Hong, Q. N., Pluye, P., Fàbregues, S., Bartlett, G., Boardman, F., Cargo, M., Dagenais, P., Gagnon, M.-P., Griffiths, F., Nicolau, B., O’Cathain, A., Rousseau, M.-C., & Vedel, I. (2018). Mixed Methods Appraisal Tool (MMAT), version 2018.Consult the manual and report criterion-level judgements with reasons for methodological appraisal.
7. O’Cathain, A., Murphy, E., & Nicholl, J. (2008). The quality of mixed methods studies in health services research. Journal of Health Services Research & Policy, 13(2), 92–98. https://doi.org/10.1258/jhsrp.2007.007074The source of GRAMMS. The EQUATOR record provides a reporting-guideline entry.
8. American Psychological Association. JARS–Mixed: Mixed methods article reporting standards [PDF].Use alongside the relevant quantitative and qualitative reporting standards to describe the strands and their integration.