Mixed methods research - blending qualitative and quantitative approaches

Part of the process of putting together your research proposal involves choosing a methodology - how you are going to collect and analyse data to answer your research questions. Like many others I began by considering whether I wanted to follow a quantitative or qualitative route, but ended up not only choosing mixed methods, but also blending the two approaches for a more flexible response to my findings.

My study concerned secondary teachers' attitudes to taking risks in their classrooms. I initially decided upon a qualitative approach as I wanted to understand what teachers considered to be risky; how they felt about risk-taking; and what drove their choices to take, or avoid risk. With so little existing research on this topic, I was looking for rich descriptions (Guba and Lincoln, 1994) and personal narratives on which to build a theory of risk-taking that might apply to teachers. At the same time, I recognised the limitations of gathering qualitative data within the timeframe of part-time PhD studies and planned to include a survey to triangulate my findings within a larger data set (Becker et al., 2012). Hence, when my research proposal was accepted, I had decided upon mixed methods.

There are different ways of using mixed methods, depending upon your circumstances. Your research questions might fall into two different categories - those that require numerical data and those targeting narrative responses. You can then apply the appropriate methodology to each individual question. I did consider this, but felt I would end up adjusting my research questions to the methods, rather than the other way round. Mixed methods can also be applied sequentially - using first one approach to gather initial data and then using a different approach to deepen understanding or triangulate findings to a larger data set. This suited my needs much better. I planned to use focus groups first to gather information on teachers' understanding and perceptions of risks. These findings would then be used to design a large-scale survey. At this point there was a fairly even weighting given to both qualitative and quantitative approaches with the first method clearly informing the second.

And then the pandemic happened.

My focus groups were cancelled and while we could have worked online, I felt the organic discussion I was hoping to facilitate would be lost. After a discussion with my supervisor, I moved to semi-structured interviews and put my mind to adapting my focus group design to work for individual conversations. I had originally put together a set of card sort activities to break the ice and prompt discussion within a group, so I decided to use them for the same purpose in the interviews. They worked brilliantly. Participants enjoyed the activities and constantly referred back to the cards throughout the discussions, which helped keep our conversations on track. What I hadn't foreseen was the impact of the quantitative data collected from the results of the card sort activities. I had originally intended to harvest data from focus group discussion as participants attempted to come to a consensus completing the activities. Now I was able to extract individual participants' solutions and record them during each interview. Qualitative and quantitative methods were being used concurrently - blending them together in my interview design.

As it happened, the early quantitative data from the card sorts suggested teachers had far more positive attitudes to risk-taking than many previous studies had suggested. I was able to adapt my initial interviews in response to these findings, which led in turn to a decision to run a second round of interviews to look more deeply at these results. In the end, qualitative methods were given the greater weighting in terms of data collection, but it was the unplanned quantitative data which gave rise to some of the most important findings in the study.

What have I learned? Mixed methods can be more than using a survey to add validity to small-scale research or using a focus group to inform survey design. There is a great deal more room for flexibility and creativity than I had first appreciated, and plenty of opportunities for gathering numerical and narrative data within one method. In my case, blending the two approaches provided important insights that might otherwise have been missed, or taken much longer to uncover.

If you are interested in finding out more about mixed methods, I would recommend reading Mixed Methods Research: A Guide to the Field by Vicki L. Plano Clark and Nataliya V. Ivankova.

Becker, S., Bryman, A. and Ferguson, H. (2012) Understanding research for social policy and social work: themes, methods and approaches. Bristol, United Kingdom: Policy Press.

Guba, G. and Lincoln, Y. (1994) ‘Competing paradigms in qualitative research’, in N. Denzin & Y. Lincoln (eds) Handbook of Qualitative Reasearch. Thousand Oaks, CA: Sage Publications.

Plano Clark, V. L. and Ivankova, N. V. (2016) Mixed methods research: a guide to the field. Los Angeles, CA: Sage.