Abstract
The title of Cooper, Glaesser, Gomm and Hammersley’s recent book may suggest, to some, a project that has already been completed. Indeed, during my research training, I was repeatedly told that to conceive of research as either qualitative or quantitative helps to perpetuate a false dichotomy between different approaches. As we know, however, much social research, however well conducted, can be classified, crudely, as proceeding in the qualitative or quantitative tradition. Research using mixed methods often aims to overcome methodological weakness in a particular method by using another to add plausibility to results. Without proper consideration of how such methods combine, and indeed whether they can be combined, single research projects can become two separate approaches to the same problem.
Part I of this book explores, in detail, some assumptions that guide common social science research approaches. The authors’ own favoured research methods are not immune from this critical treatment. They acknowledge the many difficulties researchers face when trying to identify social, and particularly causal, processes.
Instead of relying on purely philosophical and abstract discussions, the authors illuminate their treatment of some challenging theoretical concepts by the frequent use of examples. These range from small, illustrative examples of particular phenomena (akin to asking the reader to conduct a ‘thought-experiment’) to more detailed discussions based on published works. For example, Hammersley’s chapter on quantitative research into meritocracy demonstrates that not even sophisticated statistical approaches are immune from problems with interpretation, measurement and operationalisation of key concepts. Similarly, Gomm’s chapter about educational inequality stresses the need for caution when making inferences from aggregate data. By making such detailed reference to published work, the authors may be seen as attacking other social researchers or trying to discredit their work. I would, instead, suggest that the technique of explaining method through substantive examples merely mirrors what is becoming accepted in the teaching of research methods: that the subject can be too dry and difficult to communicate without some substantive context. The theme throughout this section is that no method (or group of similar methods) can be viewed as an inherently superior way to proceed in social research.
By the end of part I, it is tempting to conclude that social research is fraught, in whatever form, with so many insurmountable difficulties that the only sensible option is to give up. It might also seem, from this review, as though the authors have contributed to the very divide they seek to challenge by choosing to segment part I into chapters dealing with, in turn, methodological problems inherent in quantitative work and qualitative work. Instead, however, of merely documenting and explaining the various weaknesses in commonly used research methods, they offer a way to proceed.
In part II, the authors present an argument for moving to case-focused research. For many readers, ‘case-focused research’ suggests research with small-n samples. The authors propose that Qualitative Comparative Analysis (QCA), a case-focused method, can provide a way of conducting causal analysis on a dataset of any size. This claim is not universally accepted, even within the growing community of QCA users, and, knowing this, the authors proceed by comparing QCA with other methods used for causal analysis. Central to this approach is the explicit distinguishing of ‘empirical’ work from ‘quantitative’ work. The three chapters in this section offer detailed comparisons between QCA and other case-based methods.
Though a challenging read, Chapters 5, 6 and 7 provide some much-needed methodological comment on QCA. As it is an innovative and emerging methodological approach, it needs to undergo scrutiny and be compared with more common approaches to analysing causality if it is to be more widely accepted and used. The authors invite the reader to really ‘get under the bonnet’, so to speak, of cluster analysis, analytical induction, correlational analysis and QCA by including artificial datasets and real datasets in their examples. The use of the artificial dataset allows the authors to highlight the presence of particular methodological difficulties. Following this with some real data allows the reader to see clearly the differences between solutions found by different methods.
This book makes a thoughtful contribution to ongoing methodological debates in the social sciences and makes, sometimes, for uncomfortable reading as it highlights entrenched assumptions in common approaches. It forces the reader to scrutinise their own practices and think more deeply about whether their favoured methods can produce answers to the questions we pose in social research. Even for those who do not agree with the authors’ chosen remedy, case-focused research, Challenging the Qualitative–Quantitative Divide should inspire some self-critical methodological reflection and this can only be welcomed.