Hi Hemant,

I would argue that it is definitely the former. It is not usual to 'quantify' qualitative findings i.e. 'how many times does something occur - except perhaps with sample demographic detail. There is a analogy that I often use with students and that is 'quantitative research counts the number of sunsets - qualitative research appreciates them'. 

Mixed methods research needs to have more than one data-set. Your study will produce one qualitative data set - and so, as you suggest, it remains a 'purely' qualitative study. There is nothing to stop you identifying, in the findings or discussion section, that you noted more of one group identified an issue than the other - and identify why you might think that is the case. For me, if you wanted this to be a mixed methods study - you might follow up the individual interviews with focus group sessions - one group of one nationality; the other group the other nationality. 

I agree that your study, Hemant, is a qualitative study. And now you can consider what qualitative method/methodology is guiding your approach. Content analysis? 

Elo, S., & Kyngäs, H. (2008). The qualitative content analysis process. Journal of Advanced Nursing, 62(1), 107-115.

Graneheim, UH., & Lundman, B. (2004). Qualitative content analysis in nursing research: Concepts, procedures and measures to achieve trustworthiness. Nurse Education Today, 24, 105-112. doi:doi:10.1016/j.nedt.2003.10.001

In my experience with mixed methods, adding questionnaires would make it a mixed methods study - "stories and numbers". 

The International Journal of Qualitative Methodology is preparing a special edition on mixed methods research. 

All the best with you study! 

This is still an active topic of debate in mixed methods, where some people consider "conversion" to be a mixed methods design but others do not. I personally fall into the latter camp, because simply collecting the data one way and then analyzing it another way does not actually "mix" any methods. So, like Dean, I think that mixed methods requires both a qualitative and and a quantitative such.

And like Gwen, I would label this a content analysis.

Hi all,

It is indeed a 'can of worms'. The mixed-method/multi-method debate certainly adds to that. While I like the fact that David agrees with me on the quant/qual approach - that is  a classic distinction and not the only one. A mixed-methods approach could be purely qualitative or quantitative. It is different modes of triangulation within a single study that dictate. More than one data, investigator, theory, methodological or multi-disciplinary triangulation method, within a single study that asks pre-determined question/s from the outset, can be classified as mixed-methods. Those could all exist within just as quant or qual-only study. Beyond that, if the study is more classically a mixture of quant and qual - they are classed according to simultaneous or sequential criteria i.e. simultaneous can be quant + qual or qual + quant (both conducted at the same time) or they can be sequential - which is where either quant comes first followed by qual - or the other way around. 

Dear all, thanks a lot for your time in providing such detailed and helpful responses. I am pleased that my questions has raised such an interesting debate. In summary though it seems everyone agrees that the little quantification that I use (if I use it) will not qualify my study as mixed methods and it will still be qualitative. I do however have another question on similar lines.

Before I begin my semi-structured interviews with an open-ended interview schedule to capture the feelings and the experiences of the parents, I am planning to use a 'structured questionnaire' to capture some factual information about their current support networks in various aspects of their disabled children's life (e.g. personal care, leisure and recreation, behaviour management and so on). This questionnaire will mostly comprise of close-ended tick box type questions and will follow a strict order when administered with all the participants. Due to lack of time I might need to administer this during the same session immediately prior to the semi-structured interviews. The responses to this questionnaire, however, will not be analysed separately for their own sake and will be used primarily to provide context to the interview questions and will be merged with the finding of the interview data during analysis stage.

Based on the discussion so far, does this make my study 'mixed methods', 'multi methods' or is it still a qualitative study? Hope this question makes sense.

I'd say it depends on how you use that data. If it is simply descriptive, then I wouldn't count it as mixed methods. For example, lots of qualitative studies include a "Table One" that provides descriptive information about the participants.

Alternatively, if you do analyses on that data and produce substantive results, then you are getting closer to mixed methods. In that case, the issue would be how you "integrated" those quantitative results with your primary qualitative study

David is correct. Demographic data is usually a sub-set of data purely to describe the nature of the sample population within either a quantitative or qualitative approach. It doesn't make it mixed methods. 

Hello Helmant. While you may be able to call your research plan "mixed methods because it has both a quantitative and a qualitative piece, it wouldn't be 'good' research. As many responses state, you don't want to quantify qualitative data.

You have a very good qualitative data collection plan. Your qualitative analysis would make this good research.

You have a bad quantitative data collection plan. Sampling methods in quantitative studies, by design, are different that sampling methods fro qualitative studies. The aims of the studies are different.

If you want to understand the breadth or depth of answers to your questions, a semi-structured interview with qualitative analysis would be excellent. If you want to understand the prevalence or frequency of the answers to your questions, a selected sample of 40 respondents in 2 countries would not be good. 

If you want to do a mixed-methods study for example, do the prevalence study on a broader and more random population in your study locations and then do a follow up semi-structured interview with a convenience or key informant sample of 40 of the respondents. 

I personally don't have any objection to counting in qualitative data, so long as it is a coherent overall data analysis strategy. As I noted earlier, content analysis of this form meets those goals. A good reference on that issue is: Sandelowski et al., 2009. On quantitizing. Journal of Mixed Methods Research, 3(3), 208-222.

Note, however, that this is not the same as simply reporting things like "X% of the participants made statements such as this." Instead, content analysis requires a commitment to defining countable codes and then systematically locating those codes. 

Good responses Kent and David.

I agree David that qualitative data, under very specific conditions, might be 'counted'. That said, it's not very common and my well, for many, present as a paradoxical oxymoron. Content analysis, for many, is purely qualitative and the same as 'thematic' analysis. I think that an increasingly more common example is qualitative meta-synthesis. Of course, Hemant's study can't be one of these - as it is a single study in it's own right. Meta-synthesis, though, is still struggling to find its methodological place - and I wonder when (one day perhaps) we start to consider 'qualitizing' quantitative data. If it apparently can work one way - surely it can work the other. 

I agree with others here but would add that to take an MMR approach to this question you would have to design additional quantitative components to this study that are carried out either concurrently or consecutively and that adhere to criteria for rigorous quantitative research. For example, the initial component of this project might be a small exploratory qualitative study that includes a focus group in each locality and some stakeholder interviews. This qualitative component can help to identify the salient factors related to parenting severely disabled people. Issues such as caregiver burden, challenges with navigating the education system etc. might come up. These factors could then be incorporated into a survey which is administered to a large random sample. The key, however, is that the criteria for rigorous quantitative research are satisfied in designing and conducting this component of the study.

The close-ended questions you discussed do sound more like demographic data and are still only being administered to the small, nonrandomly selected sample that you are working with. Therefore, it would not be very rigorous quantitative data and the findings would not be generalizable in any meaningful way. A great resource on MMR is "Designing and Conducting Mixed Methods Research" edited by John W. Creswell and Vicki L. Plano Clark. Best of luck!

It sounds like a qualitative study using content analysis. It is not mixed methods because there is no quantitative component.

Thank you all for your valuable feedback and links to useful reading material. It is unlikely that I will be increasing my sample size or doing any rigorous quantification of the data - so as suggested by many of you, it will be a qualitative study using content analysis. I am pleased not only by the quality of the responses but also by the amount of interest this question generated in the resultant debate. I love ResearchGate!

Nga mihi nui (Kind regards in 'Maori'),

Hemant

Hi Hemant,

Merely including some descriptive statistics / analysis on top of a qualitative research unlikely to justify as mixed method research.  Perhaps you can include more powerful inferential statistics according to your research objective.  E.g. from the outcome of your qualitative research, there are some themes / propositions surfaced whereby you want to confirm / generalize them by conducting a quantitative research of multiple regression, ANOVA or other inferential statistics.  Or you can perform a quantitative research first e.g. via a multiple regression then deep dive with qualitative research like interview to understand those significant relationships e.g. from the angle why and how these relationships occurred. I'd done a mixed method research based on the latter approach.  You can refer to that article on:

https://www.researchgate.net/publication/259478672_Relationship_between_Team_Satisfaction_and_Project_Performance_as_Perceived_by_Project_Managers_in_Malaysia__A_Mixed_Methods_Study

All the best.

Article Relationship between Team Satisfaction and Project Performan...

Thanks Han