It is rare to have an embedded design where the quant occurs within the qual. Can you say more about your project and its goals?

First, the size of the quantitative portion of a mixed methods project depends on your goals. If you want to test hypotheses, then you will need to have a large enough sample to meet statistical standards. A power analysis with the G*Power program could help in that regard.

Second, supporting qualitative results again depends how you define that goal. For example, if you your major goal is to demonstrate that the key elements in your qualitative findings can be operationalized into measurable variables, then a relatively small pre-test sample could accomplish that.

Third, can you say more about what your "pilot test" would be intended to demonstrate?

Fourth, qualitizing quantitative data is also quite rare -- it is mostly talked about as a theoretical possibility, rather than actually done in practice. In contrast, qualitative data is frequently quantized through a content analysis that defines a relevant set of variables and then counts the frequency of those variables in the qualitative data.

First of all, it must be understood that there is no true mixed method, because what we know as mixed-method is nothing more than two types of research carried out alternately. So, each must be loyal to their respective paradigms and principles. This includes samples. Qualitative methods emphasize theoretical sampling and saturation. Choose a sample that can provide the information you need to answer the research question. Determine the sample size based on the principle of saturation, which occurs when you no longer find new information even though you have increased the sample size.

If the first stage of qualitative research produces a hypothesis that can be tested, then you can apply the exploratory sequential mixed-design model. On the other hand, if the first stage of quantitative research produces questions that require intentional explanation (for example from the actors), then the explanatory sequential mixed-design model can be chosen.

Keywords such as merging, integration and meta-inferences should be clearly visible in mixed-method research, because that is what differentiates it from other designs. How do you merge data, how do you integrate findings, and what meta-level conclusions are reached, are some of the questions that readers are also waiting for. Regards.

The key in this designs, especially with Quant under Qual, is how well quantitative data supports and explains qualitative findings. Quant sample size should be sufficient to produce meaningful, statistically reliable results without overpowering Qual's primacy and depth.

Sakban Rosidi I personally agree with you that the most effective uses of mixed methods alternate qualitative and quantitative methods in a sequential fashion. However, my bibliometric research indicates that by far the most common design in published articles that claim to use mixed methods is "convergence" where the methods are used separately and simultaneously. Then, as the name implies, the researcher brings them together during the analysis phase.

This design leads to exactly the issue that you raise: how do you integrate the findings? My main message in that regard is that one should have a solid plan for integration before beginning the research. One very useful tool in this regard is a joint display, so one could develop the basic format for a joint display as part of the initial research design.

David L Morgan

You are right. Mixed methods researchers need a personal art of integrating qualitative and quantitative findings during analysis. Of course, researchers must be aware of that when developing their research design, for example, based on the type of initial research questions. A joint display, both regarding methods and findings, and how they are integrated will really help researchers in organizing their steps and activities.

For example, based on the path analysis that I have carried out, there is one path whose coefficient is not significant (meaningless). This prompts the question, why is that? In fact, both theoretically and logically, the two variables should be linearly related. I then explored this question, by going into the field to conduct qualitative research, selecting relevant cases, so that I could explain why this happened. I believe, in research in the social and behavioral sciences, the question "why" is not simply answered by finding external causes, but also uncovering the internal reasons of the actors.

Originally I was going to include a picture, but apparently there is no facility to attach it.

Sakban Rosidi That is a very good description of an explanatory sequential design (QUAN --> qual). What worries me is poorly thought out convergent designs (QUAL + QUAN), and my research indicates they makes up at least two-thirds of the published literature in mixed methods. Sadly, there seems to be a naive assumption that having both qualitative and quantitative data is inherently a good thing, without proper attention to integration. The typical result is separate presentations of the two sets of results, as if the two studies were done without any reference to each other.

@David L Mogan, I need your guidance in my research because I have no research experience this is my first time that I'm doing research.

My interest area is to explore the teachers perception about implementing inquiry based teaching.

In my base paper there is mixed method along with constructivist theory but my concern and my university requirement is that there should be a modify theoratical model which I can run on SPSS and Smartpls.

I'll collect the data from secondary college teachers via questionnaire and interview, can you guide me how should I start and what should be my methodology.

Abila Khan It sounds like you want to use a sequential exploratory design (qual > QUAN) where you first use the qualitative data to develop a survey, and then test your hypotheses with a statistical analysis.. Is that correct?

In general, I recommend the text on mixed methods by Creswell and Plan-Clark as a good place to start. It includes a full chapter on sequential exploratory designs.

One of the best Research design to use for mixed research methodology is the concurrent triangulation design with a sequential exploratory strategy. This design combines stratified purposeful sampling for qualitative research and systematic random sampling for quantitative research.

Mustafe Hassan Dahir As far as I know, concurrent triangulation (QUAL + QUAN) and sequential exploratory ( qual --> QUAN) are two completely different designs.

I really appreciate the answers and responses given by David L Morgan.

My few comments in general, long before articles and books about mixed methods, the work of Glaser & Strauss (The discovery of grounded theory, 1967) has explored how quantitative and qualitative data can be used to build and test grounded theory. Beating the popularity of a number of previous books, now the term mixed-method design is often referred to by Creswell, both written alone (Research design: Qualitative and quantitative approaches, 1994), and with other authors (Creswell & Clark, Designing and conducting mixed methods research, 2007) .

In fact, I rather doubt novice researchers, without sufficient methodological competence and quantitative and qualitative research experience, to directly adopt mixed-methods. Especially if the research project is limited by a short time. Throughout my experience, I have been disappointed by many master's theses which were originally conceived as mixed-methods research, resulting in nothing more than two quantitative-qualitative descriptive studies, which were shallow and provided little scientific contribution.

If you still want to learn more about mixed-methods, try reading Creswell's book while listening to his explanation when giving a lecture about mixed-methods. Here is the YouTube link to the lecture:

https://www.youtube.com/watch?v=cKDeDFzMLzQ