Multiphase Mixed Methods Design

A multiphase mixed methods design refers to a researcher or research team conducting a series of interconnected quantitative and qualitative studies to comprehensively address a central issue. In this article, we will go through the different steps, when to use this design, and provide some examples of articles applying this design.

Introduction

In multiphase mixed methods design, the studies conducted are sequential designs, such that the first one informs the subsequent one. For instance, this could occur when a qualitative study guides the development of a quantitative instrument. Under this design, a single method is insufficient to fully explore an issue, but by combining methods iteratively, researchers can address various dimensions of the problem more thoroughly.

The first researchers to use this design referred to it as the "sandwich design," because it alternated quantitative and qualitative methods across three phases. Nowadays, this design combines sequential and concurrent aspects and is used in large, funded studies with numerous questions to advance an objective.

Its purpose is to address a set of related research questions and provide an overarching methodological framework to a multiyear project that needs various phases to develop a program or evaluation.

Researchers use this design when they can't fulfill a long-term program objective using a single mixed methods research design. It is also used when the researcher has experience in large-scale research and sufficient resources and funding to implement the study over multiple years. Usually, the researcher needs to be part of a team that may include practitioners in addition to experts in qualitative and quantitative methods.

Philosophical assumptions

Given that mixed methods research encompasses multiple phases, its underlying assumptions can vary depending on the specific design employed. Generally, Creswell and Plano Clark (2011) suggest that researchers often adopt pragmatism as an overarching philosophical foundation when strands are conducted concurrently. Pragmatism emphasizes practical approaches to research, focusing on what works to address the research problem without being confined to a single paradigm. For the qualitative component, researchers may align with constructivism, which posits that individuals construct meaning through their experiences and social interactions, emphasizing subjective interpretations. In contrast, the quantitative component in sequential strands often reflects postpositivism, which assumes that objective reality can be measured and understood through systematic observation and hypothesis testing, while acknowledging that absolute truth is unattainable.

It is common for different subgroups within a mixed methods study to operate under distinct philosophical assumptions, each focusing on specific aspects of the research design. This diversity in foundational perspectives enhances the study’s depth and breadth. Additionally, mixed methods designs benefit from a strong theoretical framework that guides the integration of qualitative and quantitative data, ensuring coherence across phases and providing a foundation for subsequent studies.

How to do a multiphase design?

Creating and implementing a multiphase design involves systematic planning and execution across multiple studies. This design is especially useful for addressing complex research questions requiring iterative exploration, development, and evaluation. Below is an example of how to conduct a multiphase design:

1. Define the overarching program objective. Begin by clearly identifying the broader goal your research aims to achieve. The program objective should address a significant issue or knowledge gap, providing a foundation for your entire multiphase design. This objective should guide all subsequent phases and ensure that each study contributes meaningfully to a unified purpose.
2. Plan individual studies. Divide the overarching objective into distinct, manageable studies or phases. Each phase should target a specific aspect of the broader research objective. Carefully select the type of study for each phase based on the kind of data required and the objectives of that phase.
3. Develop research questions for each phase. Each phase should have its own set of focused research questions that align with its objectives. These questions should be designed to build upon the findings of previous phases and contribute to the larger program.
4. Design study procedures. Select research methods and procedures that best address the questions of each phase. Depending on your objectives, you might choose surveys, interviews, focus groups, experiments, or observational studies. Be deliberate about choosing whether to use qualitative, quantitative, or mixed methods approaches within each phase.
5. Sequence the studies logically. Decide whether the phases will be conducted sequentially or concurrently. Sequential designs, where one study informs the next, are useful when earlier findings guide the direction of later phases. Concurrent designs, where multiple studies run simultaneously, are beneficial when time constraints or complementary methods allow parallel work.
6. Collect and analyze data for each phase. Conduct data collection in accordance with the planned methods, ensuring consistency and accuracy. Analyze the data using techniques appropriate for the phase, whether qualitative (e.g., thematic analysis) or quantitative (e.g., statistical modelling).
7. Integrate findings across phases. Once data from each phase is analyzed, integrate the findings to create a comprehensive understanding of the research problem. Use integration methods, such as triangulation, to merge qualitative and quantitative data, ensuring that the insights from different phases are synthesized effectively.
8. Iteratively refine the design. As each phase unfolds, be prepared to adapt the research design based on new insights. Refine research questions, methods, or procedures to better align with the evolving objectives of the program. For example, initial findings might reveal the need to include a new variable or address an overlooked subgroup in subsequent phases. Iterative refinement allows your research to remain flexible and responsive, ensuring that the design evolves effectively to meet the program’s goals.
9. Report results. Communicate the findings of each phase through detailed reports, ensuring clarity and transparency. Highlight how the results from each phase contribute to the broader program objective.
10. Evaluate the overall program: After the research program, assess its success in addressing the overarching objective. Consider how well the findings from each phase aligned with the program’s goals and how effectively they were integrated. Identify strengths, such as innovative methods or impactful results, and limitations, such as unforeseen challenges or gaps. This evaluation can guide future research efforts, ensuring that the lessons learned from this program inform and improve subsequent studies.

Multiphase design variants

Identifying examples can be challenging, as these designs are often published as separate projects across various journals. Variants of multiphase designs are still being explored, and efforts to classify them are in the early stages. Based on the available literature, the following variants can be observed:

• Large-scale program development and evaluation projects: These are among the most common applications of multiphase designs. They often involve federally funded research in fields like education and health services. These projects typically include phases such as exploration, program development, testing, and feasibility studies.
• Multilevel statewide studies: These studies use different methods and phases to investigate various levels within a system, such as local, state, and national levels. For instance, Teddlie and Yu (2007) highlighted multilevel projects in education, which addressed five levels: school systems, districts, schools, teachers and classrooms, and students, each requiring distinct methods.
• Single mixed methods studies combining concurrent and sequential phases: An example of this is how Fetters, Yoshioka, Greenberg, Gorenflo, and Yeo (2007) used a combined design to study Japanese women's consent for epidural anesthesia. Their approach involved a sequential phase to explore women’s perspectives using surveys and interviews, paired with a concurrent phase to study health professionals’ perspectives by collecting quantitative and qualitative data through email surveys.

Strengths

Multiphase mixed methods research designs offer several advantages:

• Flexibility in addressing interconnected research questions: The multiphase approach allows researchers to integrate mixed-method elements that can handle a complex set of interrelated research objectives effectively.
• Publication opportunities: It enables researchers to report findings from individual studies independently while still contributing to the overarching goals of the broader research or evaluation program.
• Alignment with program evaluation and development: This design is well-suited to iterative processes typically found in program evaluation and developmental research.
• Support for long-term iterative research: Researchers can utilize this approach as a framework to conduct multiple studies iteratively over extended periods, supporting comprehensive and evolving analyses.

Challenges

Even though the multiphase design is characterized by its flexibility and multifaceted nature, the same attributes present significant challenges:

• Anticipating challenges in phases: The researcher must account for the difficulties typically associated with both concurrent and sequential methods across individual and subsequent phases.
• Resource demands: Implementing multiple phases over several years requires substantial resources, time, and effort.
• Team collaboration: Effective collaboration with a research team is essential, requiring the management of team dynamics, including the addition or departure of team members throughout the project.
• Integration of studies: The researcher must establish meaningful connections between the individual studies and effectively integrate qualitative and quantitative elements within each phase.
• Practical application of findings: Given the program development focus of many multiphase designs, the researcher must consider how to translate findings into actionable materials and programs.
• Protocol updates: The researcher may need to submit updated or new protocols to the Institutional Review Board (IRB) for each phase of the research.

Examples

The National Institutes of Health (NIH) often employs multiphase designs in its behavioral and social science research initiatives, particularly in programs addressing chronic diseases such as diabetes and heart disease. These large-scale projects start with exploratory qualitative studies to understand patient behaviors, beliefs, and challenges in managing their conditions. For example, interviews and focus groups may uncover barriers to adopting healthier lifestyles or adhering to medication regimens. Quantitative studies then follow, testing specific interventions, such as diet or exercise programs, on diverse populations. Finally, large-scale evaluations assess the effectiveness and scalability of these interventions across different settings, providing actionable insights for public health policies.

Educational research often uses multiphase designs to explore systemic issues, as demonstrated by Teddlie and Yu (2007). Their work examined inequities across different levels of the educational system, including school districts, individual schools, classrooms, teachers, and students. The design began with qualitative studies to capture the perceptions and experiences of teachers and students. Quantitative surveys followed, measuring disparities in resources and outcomes across districts. Mixed methods were used to integrate these findings, resulting in actionable recommendations to reduce inequities and improve educational policies at multiple levels.

The World Health Organization (WHO) frequently uses multiphase designs in community health programs to address disparities in underserved regions. These initiatives typically start with qualitative assessments to identify specific community health needs and challenges, such as limited access to clean water or healthcare services. The second phase uses quantitative surveys to map the extent of these issues and identify at-risk populations. Interventions, such as vaccination campaigns or health education programs, are then developed and pilot-tested. The final phase evaluates the interventions’ effectiveness and scalability, ensuring they can be adapted to different contexts and communities.

Conclusion

Multiphase mixed methods research design stands out by integrating qualitative and quantitative approaches across multiple, interconnected phases. Unlike simpler mixed methods designs—such as convergent, explanatory, or exploratory sequential designs that focus on a single phase or specific sequence—multiphase designs address complex research questions through a series of studies or phases. This structure allows researchers to explore, test, and refine concepts iteratively, with each phase informing the next.

The key advantage of this design is its flexibility to tackle diverse objectives within a cohesive framework. While it offers richer insights and stronger validation of findings, it also demands careful planning to manage its complexity and ensure integration across phases. DDespite these challenges, multiphase mixed methods design is powerful for generating comprehensive, actionable knowledge that stretches beyond the capabilities of any single mixed methods research study.

References

1. Creswell, J. W., & Plano Clark, V. L. (2017). Designing and conducting mixed methods research (3rd ed.). SAGE Publications.
2. Fetters, M. D., Yoshioka, T., Greenberg, G. M., Gorenflo, D. W., & Yeo, S. (2007). Advance consent for epidural anesthesia: A mixed methods study of Japanese women's preferences. Birth, 34(3), 220–229.
3. Teddlie, C., & Yu, F. (2007). Mixed methods sampling: A typology with examples. Journal of Mixed Methods Research, 1(1), 77–100.