Mixed methods research in education : some challenges and possibilities

Publication Summary

It was with great pleasure that I agreed to address the 2012 conference on mixed methods as part of UTDANNING2020 Research Council of Norway. My thesis was that what is usually called ‘mixed methods’ research in education is really just research in education. It is relatively easy to conduct, with many possibilities and few real-life challenges or barriers. What this paper tries to do is convey is part of why this is so. There are of course many different methods of investigation that could be said to be ‘mixed’ in any one study – interviews with documentary analysis, or multiple regression with inferential statistics, for example (Symonds and Gorard, 2010). However, for the purpose of this brief paper, the mixture is assumed to refer to those methods that have traditionally labelled ‘qualitative’ and ‘quantitative’. For some reason, social scientists have long separated any data that involves counting or measuring from all data that involves anything else – text, conversations, observations, smells, drawings, acting, music and so on. I have no idea why. But such social scientists say that these two groups – numbers and everything else – are incommensurable, and require a completely different logic to use, and have unmatched criteria for judging research quality, and many other purported differences. Then, just to confuse things, some social scientists say that we can and should mix these forms of data – and that presumably they are not commensurable in combination, only in isolation if that makes any sense at all. It is no wonder that new researchers are confused, and that the potential users of social science evidence just ignore us. We live in a kind of la-la land.