Triangulation involves uses two different data sets to find out if they give you the same answer to your research question. Steps 2 and 3 sound like they could have accomplished that, except that you used the same method both times. And if you asked the same questions both times, then this would be more like replication than triangulation, and successful replication would be clear evidence for the credibility of your findings.

To me, taking your results to an expert panel for evaluation sounds more like member checking -- which would also be a means for strengthening the credibility of your results.

Triangulation involves uses two different data sets to find out if they give you the same answer to your research question. Steps 2 and 3 sound like they could have accomplished that, except that you used the same method both times. And if you asked the same questions both times, then this would be more like replication than triangulation, and successful replication would be clear evidence for the credibility of your findings.

To me, taking your results to an expert panel for evaluation sounds more like member checking -- which would also be a means for strengthening the credibility of your results.

There are four types of triangulation; methodological, researcher, data and theory.  

Please read to identify yours.

http://www.pm.lth.se/fileadmin/_migrated/content_uploads/Jick_1979__Mixing_qualitative_and_quantitative_methods_-_Triangulation_in_action.pdf

Taking your results to the expert panel is; cross checking to vaildate. It is a good back up for validation in a qualitative study. Please do not forget to report interreliability, too.

Have a great week.

Hello Raman,

David and Nuray above have given you some interesting and valuable things to think about. I have a somewhat different angle to take about what is happening in your research.  That angle is not necessarily correct, but it might be helpful if I share it with you.

In essence, I see it as a series of phases, all but one of them "legitimate", but none of which seems to be concerned with triangulation.  In the first phase, it seems you were identifying where you wanted to go. This is good foundational work.  In the second phase, you worked with what you'd established in Phase 1 - and it seems that you could have engaged in some triangulation but didn't choose to. To me, that's quite OK. You were analysing the data in as effective a way as you could  

In Phase 3, I think you were really establishing, as you suspect, generalizability, of your model.  ( I can't see it as being triangulation, but maybe I am overlooking something. And to me it doesn't look like replication, at least in the full sense, because you deliberately chose a different kind of sample.)

In Phase 4 (Step 3B),  I'm not sure what's really going on.  It sounds as if you might have gone in a circle, however, and I'm not sure what the value of this step is therefore. Maybe I have not understood something, but if you created a structure from Step 1, then obtained data based on that structure in Step 2, it's not surprising that the structure in Step 2 matched that of Step 1.

In Phase 5 (Step 4), it does sound pretty much as if you are engaging in something like member checking (not triangulation) if you indeed go back to the participants from Phase 3 to see if your results ring true to them. If you had already done member checking back higher up, and had positive results from that, I wonder whether this step is really necessary. 

Although triangulation is one means of ensuring the validity of your results, there might be other ways, such as member checking, that you need to rely on. Of course, it might depend on how much time and how many resources you have at your disposal.

I hope all of the above is helpful.  Every good wish with your research.

Robert.

Dear Mr. Raman K. Attr,

Please reference this book:

I agree with what David Morgan has commented, and also Robert Trevethan.  What struck me when I read your interesting question was the absence of a report of either the research question, or the paradigm.

For me, concepts like 'triangulation' are only meaningful in the context of what is being sought, and how you conceptualise the status of the knowledge derived from your methods. 

I encourage students who are using Interpretive Phenomenological Analysis for example, to think more deeply about what they really mean by triangulation and what status it might have.

For my own interpretive research using a complex realist paradigm to study professional practice interactions, I think more about 'enriching' than triangulating, and about different methods pointing me towards less accessible phenomena, (eg, 'why is this person not talking now?') rather than verifying that a particular data point is 'right'.

Further to the above knowledgeable views and comments, may I suggest you refer to below paper, specifically Olsen’s (2004, p. 13) comment: triangulated research may run the risk of taking on too many unfocused questions all at once unless…… 

• Olsen, W., (2004) Triangulation in social research: Qualitative and quantitative methods can really be mixed, in Haralambos, M. and Holborn, M. (Eds.), Developments in Sociology, Causeway Press, Ormskirk, UK,  pp. 1-30.

Dear Professors,

Here is "Triangulation in social research: Qualitative and quantitative methods can really be mixed" in file.

Best regards.

Dear colleague, 

You can also have a look here: 

Blanco, T., Berbegal, A., Blasco, R., & Casas, R. (2016). Xassess: crossdisciplinary framework in user-centred design of assistive products. Journal of Engineering Design, 27(9), 636-664

(PDF attached)

In this paper we present Xassess, a new methodological framework that interweaves different phases and types of assessments, with a set of tools to be used individually or conjointly. It is based on evaluation and decision making along the project stages and it can be applied to many design or project development realms (not only assistive products). We talk about mixed methods, mentioning the differences between combination, complementation and triangulation, and offering some applied examples that I hope you will find useful. 

Best regards

Hi, Ramen,

I am answering just by reading your questions without reading the answers from other researchers, so please bear with me if there are duplications.

Step 2; I think it is okay to consider it as "data triangulation".

Step 3: I will consider it as a strategy to establish "generalizability". Because, by and large, you are still using interviews. so the methodology is more or less the same. 

Step 3b. I don't see you can call it either date or methodological triangulation. It is jsut that the data are matching. I understand triangulation is different means (purposeful/intended or prior to research, means) that you use to gather data from  planned target sample groups.    

Step 4: if you add focus group or use Delphi, then I will think it is a methodological triangulation of data that can be compared and analysed vis-a-vis other data. 

But you also need to clarify if you want to use the same sample group (or at least some members) for the focus group. Then, it will be a data triangulation, becomes group members are the same but in different atmosphere you can check if the data you received from the same group will differ.  

I know Delphi, Try to goolge a Nine cross-country research on civic education of social studies by John Cogan, David Grossman, Somwang...et. als. That was around 1998 -2001.  They use Delphi, in their cross-country research.  otherwise, try Patricia Kubow PhD thesis for Civic education, THat is a thesis from University of Minnesota. I think that used Delphi too.     

Hope these help. 

First, much of your work from phase 1 on, seems targeted towards answering essentially different research question rather than increasing the validity of the answer to a single question, which makes me question whether 'triangulation' is the correct term here. See...

Bryman, Alan. "Why do researchers integrate/combine/mesh/blend/mix/merge/fuse quantitative and qualitative research." Advances in mixed methods research (2008): 87-100.

Second, your approach triangulates (- in a loose sense) between orientation, exploration and testing, validation as well as between use, relevance, users and reuse,rigor, experts. These distinctions are well described in the DOT-framework (see attached), so this appears to be an applicable framework.

Conference Paper Design Patterns for Mixed-Method Research in HCI