Designing and Conducting Mixed Methods Research
A story. In the beginning, there was Quantitative Research. A few years later, there was Qualitative Research. And then Mixed Methods was born, as a separate child of these parents. Let's all welcome Mixed Methods. As it happens, I am very committed to the idea of mixed methods research, dislike the idea of any kind of hierarchy of methods (as opposed to an hierarchy of evidence), and firmly believe that all public health researchers should be conversant with the basic methodologies and methods of both schools as well as a number of others, including systematic reviews, evaluation, documentary research and so on. Each of these has its own usually complex set of methods and practices. A book addressing mixed methods, therefore, should be a welcome addition to the teaching library. Designing and Conducting Mixed Methods Research is organised into 10 chapters that are intended to be logically sequenced, from a set introducing mixed methods as a research style, through design, data analysis and writing up to questions and future directions for mixed methods. However, the chapter titles do not necessarily reflect the contents, and some important foundations of research theory are missing. For example, ontology and epistemology are included but scantily dealt with, and while specific qualitative methods do receive some description, quantitative methods seem to be thought of as a homogenous bolus. Many useful data collection techniques, such as Delphi and Most Significant Change, are also missing in action. The book includes several frameworks and tables, some useful and some very strange and impenetrable. The book contains many unreferenced inconsistencies and truisms, so that it is somewhat idiosyncratic to read. For example, chapter 1 is about understanding mixed methods, and begins: " Work on this book began almost a decade ago when we started writing about mixed methods research at the time that qualitative research had achieved legitimacy and writers were advocating for its use in the social and human sciences " . Sorry? Date check, this is published in 2007. Do the authors truly think that it was 1997 before qualitative research gained academic legitimacy? The authors introduce us to four types of mixed methods designs: embedded, exploratory, triangulated, and explanatory, with some detail about each of the procedures and designs for each. However, the book has some big gaps in basic research techniques. For example, looking up the word 'sample size' in the index leads to the sample size question. Answer? The sample size for qualitative arm of the study (preferably purposefully selected) will be smaller than for the quantitative (preferably randomly selected) arm. And if you increase the number in the qualitative arm to match the quantitative arm you will lose detail, and according to the authors this effect is inevitable. But why? Another example: running quantitative and qualitative arms concurrently with the same participants has the potential to bias data, but again I cannot understand the logic of the argument against this practice; indeed, in most quantitative study there is at least one of those responses that invites some sort of thoughtful comment, which I have always assumed to be a perfectly internally valid reflection of the writer's thoughts. Analysis? Well, you need to know how to do it appropriately! Well I never! The book is generally poorly referenced and some excellent examples of mixed methods research are notable for their absence. On the other hand, the authors reference their own work extensively (overall about one-tenth of the 10 pages of references), giving the book something of a house-of-cards feel; disagree with an argument and the whole section collapses. Roughly half of the index also consists of the names of writers to whose work the authors refer (who are largely American, so there are notable gaps in it). After much trying, I cannot see how mixed methods as an independent research design differs from the many research projects that have designed into them from the beginning both qualitative and quantitative arms. In short, despite the inference in the title, a student could not use this book as a stand-alone manual, and while it contains some useful ideas it would not be a particularly useful public health research methods course book.