Hello!!
I am using a mixed-method research design. And this is my first ever experience with MMR approach. So please, I need guidance to understand it more clearly.
I intend to conduct the data analysis in the following steps;
The number of surveys four.
Step 1) Survey A compared and contrasted with Survey B, C and D.
All four surveys contain the same format and sets of questions, quantitative questions with interlinked open-ended questions. (Predominantly Quant)
Step 2) Survey II compared with Survey A, B, C and D.
Survey II contains open-ended questions and which are going to be compared with interlinked questions of Survey A, B, C and D. the purpose is to determine the gaps and interdependency between Survey II and Survey A, B, C and D.
My question is;
How will you describe and define this type of analysis process, concurrent embedded or Triangulation design?
Many thanks for the support.
Kind Regards,
Mehtab
It sounds like you could be using a sequential explanatory design (QUAN --> qual). In that case you would begin by producing quantitative results in the first segment of your study, and then use the follow-up qualitative methods to help you understand those results.
Many thanks, David L Morgan for the reply.
I have collected the data concurrently. And there are two different audience samples. I am carrying out analysis concurrently as well.
I have just split the analysis process into two because of the two different audiences involved.
I have used a questionnaire with the interlinked open-ended question for step 1 audiences and for step 2 audiences I have used open-ended questionnaire.
What confuses me is the next stage of convergent and integration of the data.
I am comparing the same type of data (so is it called concurrent embedded design?) and also I/m comparing two different data sets against each other (so is it a triangulation?) Or can I use both the terms together?
Thanks!
It is quite common to have difficulty with integrating data that were collected in two separate studies, which is commonly known as a parallel convergent design. But the name of this approach already points to trouble, because things that are parallel do not converge and things that converge are not parallel.
Your design is not embedded because one study was not contained within the other. It is probably but not necessarily concurrent, because a concurrent design depends more how the data were used, rather than simply when they were collected. So, if the questions in your data collection were carefully connected to each other across the two studies, you could still use a QUANT --> qual design, even if the data were collected concurrently.
As for comparing the data across the two studies, the term triangulation is no longer preferred for this because it has too many different meanings. In particular, the editors of the Journal of Mixed Methods Research have stated that they will not longer review papers that claim to use triangulation as a design. To some extent, this is just a matter of terminology, but if you are comparing a qualitative and a quantitative result for the same research question, then I would call a case of convergence. In other words, do your two sets of results converge on the same answer?
Many thanks, again David L Morgan for your reply.
So can I go on comparing the datasets without having merged or converged them? Please advice!
I there any literature you could suggest that support this similar kind of analysis in MMR, please?
Thanks!
- Badges
- Science topic
- Similar topics
- Statistics
- Survey