Are you pondering your research design and have been advised to look into mixed methods research?
Then you’ve come to the right place at the right time.
In the next few minutes, I will provide you with the basics of the mixed methods approach. You’ll learn when it makes sense to combine qualitative and quantitative data and analysis elements. Additionally, you’ll become familiar with the most common methodologies and some helpful foundational texts.
This way, you can quickly decide whether mixed methods are suitable for your work and where to continue your reading on this topic.
The Philosophical Backstory of Mixed Methods (Highly Simplified)
To understand the concept of mixed methods, it’s worth taking a brief look into the philosophy of science.
Epistemology refers to theories of knowledge and describes how researchers can convert reality to knowledge. In short: How is knowledge created?
In today’s research landscape, we can broadly distinguish between two prevailing epistemologies: Positivism and Interpretivism.
These camps have long debated which paradigm is the true one. In specific research disciplines like business studies, sociology, or political science, both camps can be represented.
Some disciplines are so dominated by one camp that the other is rarely recognized. For example, natural sciences are truly positivist – which means that there is only one objective natural world out there, which can be best represented by numbers and maths.
Some social sciences like psychology adopt this view and others are somewhere in the middle. Here, different philosophical stands are accepted and methodologies are followed. These methodologies are classically either qualitative (on the side of interpretivists) or quantitative (on the side of positivists).
Since the social sciences have become much more pluralistic, the combinations of these methodologies has become more common and the advantages of “the other side” are appreciated.
Mixed methods were born!
Definition of Mixed Methods
To speak of mixed methods, a study design must include both qualitative and quantitative methods.
This means that the study design is intentionally developed with this combination in mind, and the research question can only be answered through the combination.
The choice of method or combination should always be closely linked to the research question, research goal, and the context of the research.
The first question you should ask yourself is:
What added value do mixed methods offer compared to a study design with only qualitative or only quantitative methods?
To help you with this, here are some advantages of mixed methods that you can use to justify your study design.
Advantages of Mixed Methods
#1 Mixed Methods can simultaneously answer open and closed research questions
What does that mean?
Through the qualitative part of the study design, you can explore and answer an open research question (e.g., How does an infodemic, a spread of misinformation, propagate on social media?).
With the quantitative part, you can test specific hypotheses (e.g., a warning label indicating unverified third-party information reduces the spread of an infodemic), which helps answer a closed research question.
Additionally, qualitative methods and open research questions often aim at developing new theories (exploratory research), while quantitative methods and closed research questions typically test existing theories (confirmatory research).
#2 Mixed Methods can offset the weaknesses of each method
For example: Qualitative interviews can add depth to a scientific study, but often only a small sample of experts can be interviewed.
If you additionally send a quantitative online survey based on your interview findings to many more people, you achieve the breadth that a purely qualitative study couldn’t provide. This could, for example, increase the statistical generalizability of your findings.
The argument works in reverse as well.
#3 Mixed Methods can yield contradictory results
This might initially sound like a disadvantage. However, it’s not. Contrasting results from both approaches can provide a deeper understanding of the phenomenon and highlight the limitations of each single methodology. This leads to more discussion points and more nuanced results.
Should mixed methods always be used then?
No, of course not. It all depends on the research question, research goal, and context. If your work aims to test existing theory, incorporating qualitative elements might not make much sense.
If you are at the early stages and exploring a topic scarcely covered in existing literature, you might focus on qualitative exploratory research only.
Implementing Mixed Methods
When implementing a mixed methods study, you need to be clear about why you are doing it. This will also determine the study design and the chronology of implementation. Here are the three most common variants of mixed methods designs.
#1 Complementation
In this variant, qualitative and quantitative elements are equally prioritized and are intended to provide complementary results on the investigated phenomenon.
Order: Quantitative (50%), then qualitative (50%). Or vice versa.
#2 Completion
In this variant, one of the methods is prioritized and subsequently supported by the other to ensure the phenomenon is fully covered.
Order: Quantitative (90%), then qualitative (10%). Or vice versa.
#3 Sequential Designs
The most common mixed methods studies follow a seuqential design. This means you complete one method first, and then start with the other.
Explorative sequential design
A qualitative study is used to develop constructs or hypotheses, which are then tested with a quantitative study.
Order: Qualitative (exploration), then quantitative (testing).
Explanatory sequential design
First, A quantitative study is used to test hypothesis. Second, a qualitative study follows to understand “why” these results occured.
Order: Quantitative (explanation), then qualitative (understanding).
#4 Parallel Designs
The alternative to a sequential design is a parallel mixed methods design. Here, you conduct multiple methods at the same time.
In this case, the second study does not build on the results of the first, but instead, the results of both studies are compared and contrasted once they are completed.
Validating Results
For mixed methods research, validating the results is an important quality criterion.
For quantitative data, there are well-standardized calculations (e.g., Cronbach’s Alpha) that can validate constructs using SPSS or similar programs.
For qualitative data, the validation process is much softer, and there is less consensus on the standard. Leung (2015) suggests these five criteria:
- Does the research question align with the analysis results?
- Is the chosen methodology suitable for answering the research question?
- Does the research design match the methodology?
- Is the sample appropriate for studying the phenomenon?
- Do the results and conclusions fit the sample and the research context?
You basically apply the same techniques to ensure validity and reliability as you would for just one method (Venkatesh et al., 2013). Now you do it for two. This is why mixed methods often means more workload, but in the end, you also have a more valuable study.
What is the difference between mixed methods and Triangulation?
In mixed methods, your qualitative and quantitative parts are typically treated equally. If one method is only worth 10% or so, then you can speak of triangulation.
The purpose of triangulation is to provide additional perspectives to the object under study, for example by adding additional researchers, another theoretical perspective, or addition data that is all different from the main study.
Mixed methods is highly respected because it has some aspects of triangulation already built-in. However, you can also perform triangulation without using quantitative and qualitative elements. Therefore, mixed methods and triangulation are related concepts but not the same!
What to read next?
If you now want to deepen your understanding, I suggest you get your hands on a copy of Creswell and Clark’s “Designing and Conducting Mixed Methods Research” (2010).
It is one of the standard methods book on this topic and it is applicable for any social science discipline!