Giorgio,

How come your question is almost exactly the same as one asked an hour before yours was posted on ResearchGate?

Dean

Ahh - just noticed. You are part of try same research team. Always best to just send out a singular team request - rather than multiple request from the same team - on any forum.

Are you asking for methods to insure the rigor of your analysis and the dependability of your findings?

Good morning Giorgio,

I believe that everything depends on what you are investigating, what are your research goals and what kind of design you have developed to carry out the mixed research. I think that it is totally valid for you to construct your own methodological justification, looking for a correspondence between your objectives, the nature of the research object and the operationalization of the variables. For example, if the methodological design is quantitative-qualitative, that is, it is fundamentally quantitative, the representativeness of the sample, the degree of generalization of the results, and the adjustment of the process to what was established from the beginning of the investigation are basic to demonstrate scientific validity and rigor. It could be very useful to consult the text of Roberto Hernández Sampieri, Carlos Fernández Collado and Pilar Baptista Lucio, specifically the chapter on mixed research. Please, if you need more information, do not hesitate to contact me.

Best regards.

Adding qualitative research to quantitative data makes it more valid because qualitative data makes up for the limitations in quantitative data and vice versa. If you can get comparative evidence from many different sources the more well-rounded and more representative your data/results will be. Quantitative is overrated and incomplete. Quantitative data can only tell you what and how many, actual inferences cannot be made without qualitative research because it can tell you why and how .

En el caso de la investigación cualitativa (especialmente en Antropología), el asunto es que la información es muy variable, porque, generalmente, se obtienen datos diversos en cada caso; en una investigación cualitativa es muy común no contar previamente con variables independientes y dependientes. Las variables van saliendo durante la investigación.

La pregunta entonces es ¿cómo combinar un análisis cualitativo de casos particulares y diversos con el análisis estadístico que si considera variables independientes y dependientes similares para todos los casos estudiados?

Quisiera más información sobre los tres conjuntos de resultados obtenidos para tener una pista sobre la pregunta que les preocupa.

How would you suggest to validate results from mixed method research?

Firstly, what is your goal using the mixed method research? If the goal is to represent various variables/ parameters/ subsystems to be a system, you can compare between the system required outputs and the system realized outputs. If there are some gaps, you should evaluate the gaps or you can reset some variables which build the system. Do the test repetitively until the gaps are minimal as you want.

Secondly, if the mixed method research is various methods applied (qualitative or quantitative) to validate an existing model, for example, just do the test as in the first step above, then compare among the various tests, which one is delivering the least gaps. Choose the minimal gaps. Here, the validation is the test method by comparing the model results and the real results of a system.

Easy really. Ask: have I answered my research question / achieved my research purpose by implementing a transparent research process which includes aligning methods of data collection and analysis with question / purpose?

This applies whether your research is mono-quantitative / -qualitative or mixed methods.

There is of course, the ongoing (interminable?) debate about mixing quantitative with qualitative research which sometimes seems to result in muddled thinking on the part of some researchers; for instance, in my field of substance misuse research, the late John Booth Davies rails against mixing methods:

"...the notion that qualitative and quantitative methods, separated by veritable philosophical chasms, can simply be cobbled together side by side within the same research exercise simply will not do."

(Drugspeak, 1997, P. xi, para. 3.),

and then goes on to do precisely that in his team's development of the Functional Discursive Model of 'drugspeak' which is obviously a product of mixed methods since it gauges the authenticity of drugspeak as an average of utterances over a range of scenarios.

Perhaps the resolution of the '2000' year-old debate' is for researchers to simply implement their methods with the philosophical perspectives suited to each---inclusive of each in mixed methods research; I don't really think that there is a contradiction in this.

All that really matters is how you arrive at a transparent explication of your research findings.

A standard technique to validate results is to member check. Can you take the results back to the people who were interviewed? They could give insight as to which is more accurate.