Mixed Methods Research: Integrating Qualitative and Quantitative Findings

In modern research, social, behavioral, and economic phenomena are increasingly complex and multi-layered. Relying on a single method—either qualitative or quantitative—often fails to capture the full nature of real-world problems.

That’s why Mixed Methods Research has emerged as a rigorous, practical approach: it combines the exploratory depth of qualitative data with the generalizability and statistical power of quantitative data. Designing and Conducting Mixed Methods Research (Creswell & Plano Clark, 2018) is widely cited for framing mixed methods as an intentional way to achieve a more complete understanding than using a single method alone.

In Vietnam, mixed methods are increasingly used across social sciences, education, healthcare, marketing, and organizational behavior—especially when researchers need both:

• measurable outcomes (e.g., learning gains, performance indicators, adoption rates), and

• human explanations (e.g., motivations, barriers, lived experience, context).

However, mixed methods is not “doing qual + quant in one paper.” The defining feature is integration—the intentional way you connect, merge, or embed datasets across design, data collection, analysis, and interpretation. This “integration-first” mindset is strongly emphasized in Achieving Integration in Mixed Methods Designs (Fetters, Curry & Creswell, 2013).

This guide walks you through fundamentals, core designs, a step-by-step workflow, real-world applications, reporting standards, and a tool stack (NVivo, XLSTAT, SmartPLS, EViews, Citavi) that supports mixed methods work end-to-end.

What Is Mixed Methods Research?

Mixed Methods Research is an approach that collects, analyzes, and integrates both qualitative and quantitative data within a single study (or a closely linked program of studies) to generate a more comprehensive understanding of a phenomenon.

A key point from Creswell & Plano Clark’s mixed methods textbook is that value comes from combining methods intentionally—not running two separate projects that never connect.

The key idea: integration (not duplication)

If qualitative and quantitative components don’t “talk to each other,” you usually end up with:

• two parallel result sections, and

• a conclusion that repeats both without a unified claim.

High-quality mixed methods does the opposite: it turns two datasets into one stronger inference through planned integration—exactly what Fetters et al. (2013) explains with practical integration principles.

Why Choose Mixed Methods Research?

1 Core advantages

Stronger validity through triangulation
By checking patterns across different data types, you reduce method bias and strengthen confidence in your claims—especially for applied research where context matters.

A more complete and nuanced explanation

• Quantitative helps answer: How much? How many? To what extent?

• Qualitative helps answer: Why? How? Under what conditions?
  Together, you can measure outcomes and explain mechanisms.

Highly relevant for applied research
This is especially useful in education, business, marketing, policy, and healthcare—where human behaviors and measurable outcomes interact.

Supports theory development and testing

• Qualitative can generate concepts, hypotheses, or measurement items.

• Quantitative can test relationships, structure, and generalizability.

 

2 When to use mixed methods

Mixed methods is a strong fit when:

• your research includes both exploratory and confirmatory questions,

• you need to explain surprising statistical results with qualitative data,

• you want to develop a scale/model from qualitative insights and validate it statistically, or

• stakeholders require both “numbers” and “reasons” for decision-making.

Common Mixed Methods Research Designs

Mixed methods designs differ by timing (parallel vs sequential), priority (which strand leads), and how integration happens. Standard design typologies are summarized clearly in Creswell & Plano Clark (2018) and applied integration examples are discussed in Fetters et al. (2013).

1 Convergent (Parallel) Design

What it is: collect qualitative and quantitative data at the same time, analyze separately, then compare and merge results.

When it works best:

• you can collect both datasets within the same time window,

• you want corroboration (do the stories match the statistics?),

• you need a balanced view quickly.

Example (Vietnam marketing):
Survey 400 customers on satisfaction + 12 in-depth interviews to uncover why certain segments rate low despite high usage.

Integration move: comparison + a joint display (a table/figure combining results).

2 Exploratory Sequential Design

What it is: start with qualitative data, then build instruments/models for quantitative testing.

When it works best:

• the phenomenon is new,

• existing scales don’t fit the local context,

• you need to develop a framework or hypotheses first.

Example (education):
Interview students to define “digital learning experience” → generate items → survey and validate a scale.

This design is often thesis-friendly because it gives a clear storyline: discover → build → test.

3 Explanatory Sequential Design

What it is: start with quantitative results, then use qualitative methods to explain statistical patterns, outliers, or unexpected findings.

When it works best:

• your survey shows strange subgroup differences,

• your model “works” statistically but lacks mechanism,

• you need stakeholder-friendly explanations.

 

4 Multiphase (Complex) Designs

Used in long-term or large-scale research programs with multiple phases and methods (e.g., program evaluation, product/service innovation cycles).

Mixed Methods Research: Step-by-Step Workflow

 

Step 1: Define your integration purpose early

Before choosing a design, clarify what integration must achieve:

• Confirmation (verify patterns across datasets)

• Explanation (explain statistical results with qualitative insights)

• Development (use one strand to build tools/models for the other)

• Expansion (examine different dimensions of a phenomenon)

Step 2: Write research questions that require mixing

A common mistake is writing only quant-style questions and later “adding interviews.”

Instead, structure questions so each strand has a role:

• Quant question: To what extent does X predict Y across groups?

• Qual question: How do participants describe the mechanism behind X → Y?

• Mixed question: How do qualitative themes explain or refine the statistical relationship?

Step 3: Choose a design you can execute with your timeline

A thesis-friendly rule:

• short timeline + access to both data sources → convergent

• new topic + need scale/framework → exploratory sequential

• existing dataset + need explanation → explanatory sequential

Step 4: Build a sampling plan that supports integration

Your samples must align with your integration logic:

• Convergent: same population, comparable subgroups

• Explanatory sequential: use quant results to select interviewees (e.g., extreme/typical cases)

• Exploratory sequential: ensure qualitative diversity so later survey items represent the domain

Step 5: Collect data with “integration hooks”

Design features that make integration easier:

• link interview questions to survey constructs,

• include an open-ended survey question (embedded qual),

• collect a shared identifier (segment/cohort/role) to compare patterns.

Step 6: Analyze with the right tools

For qualitative data, NVivo supports organizing and analyzing unstructured data (documents, transcripts, audio/video) to identify patterns and themes.

For quantitative data, common options include:

• XLSTAT for statistical analysis within Excel

• SmartPLS for SEM/PLS workflows

• EViews for econometrics, time-series, forecasting, and modeling.

For literature and knowledge organization, Citavi supports reference management plus structured knowledge organization.

Step 7: Integrate results (the heart of mixed methods)

Integration can happen through:

• Merging (bring results together at interpretation)

• Connecting (use results of one phase to shape sampling/instruments of the next)

• Embedding (nest one dataset within the other)

A highly recommended strategy is using joint displays—tables/figures that place qualitative themes next to quantitative results to show how one explains or expands the other. This is discussed directly in Fetters et al. (2013).

Step 8: Report mixed methods properly (don’t hide the integration)

Reporting standards increasingly expect you to clearly state:

• why mixing is justified,

• which design you used (timing/priority),

• where integration occurred and how it occurred.

A widely used reference is APA JARS–Mixed.

Real-World Applications of Mixed Methods Research

Education

Evaluate a new teaching model by combining:

• Quantitative: grades, completion rates, engagement metrics

• Qualitative: interviews with students and instructors to explain what worked and why

Business & Marketing

Consumer behavior research:

• Surveys for satisfaction/NPS/usage

• Interviews or focus groups to explore motivations and barriers

Healthcare & Social Sciences

Assessing community health programs:

• Statistical indicators (recovery, access, utilization)

• Patient stories and contextual insights (barriers, stigma, resource constraints)

Essential Tools for Mixed Methods Data Analysis

• Qualitative: NVivo Lumivero

• Quantitative: XLSTAT, SmartPLS, EViews Lumivero+2SmartPLS+2

• Literature & knowledge organization: Citavi Lumivero

Quality, Rigor, and Common Pitfalls

Pitfall 1: “Two studies in one” (no integration)

Fix: define an integration purpose + plan a joint display from day one, as recommended in Fetters et al. (2013).

Pitfall 2: Weak sampling logic

Fix: align sampling with the design (e.g., use quantitative results to select interview cases in explanatory sequential).

Pitfall 3: Reporting is vague (“we did both methods”)

Fix: explicitly state design, priority, sequence, and integration points following APA JARS–Mixed.

Reporting checklists you can use

• GRAMMS (Good Reporting of A Mixed Methods Study) EQUATOR entry

• APA JARS–Mixed

Appraising mixed methods quality (especially for reviews)

If you’re doing a systematic mixed-studies review (or want a structured quality check), MMAT (2018) criteria manual (PDF) is a widely used appraisal tool covering qualitative, quantitative, and mixed methods components.

Conclusion: Mixed Methods Research for Modern Evidence

Mixed Methods Research bridges scientific rigor and real-world complexity. When done correctly, it helps you:

• extract deeper meaning from complex, multi-layered problems,

• strengthen validity through cross-evidence,

• build and test theory within one coherent study,

• produce findings that are both publishable and actionable.

The practical takeaway: mixed methods succeeds when integration is planned, not added at the end. Use a clear design, build integration hooks into sampling and instruments, analyze with appropriate tools, and report using recognized guidelines.

Mobilio Support (Vietnam)

Mobilio supports researchers end-to-end with official academic tool licensing and practical onboarding for NVivo, XLSTAT, SmartPLS, EViews, and Citavi from methodological design to mixed data analysis.