Friday 20th April 2018, 13:00 – 14:00, BEIS, 1 Victoria Street, London
Presenter: Professor Wendy Olsen, University of Manchester
In development studies, one wants an evaluation team to face up to the challenge of combining surveys with semi-structured interview data. In this seminar I explain and show how these linkages are made with concrete data about villages in Bangladesh and India. The problem is that complex data can overwhelm the interpreting team. The solutions I offer here are methodologically sound.
In a series of past publications I have investigated aspects of indebtedness and wealth in India, Sri Lanka and Bangladesh. I used case-studies, quotations and interpretation, and hence discourse analysis. We can use Qualitative Comparative Analysis (QCA) to bring part of the explanatory analysis to fruition. This is a systematic method which helps us interpret patterns in the data. Using QCA methods, we augment the survey using evidence at multiple levels. We can insert national level evidence if we are making a comparative study of households; or we can add personal evidence from interviews based on localised studies to contrast the interactive effect of social classes and ethnic groups; or we can link multiple rounds (repeat visit) survey data.
In this talk, I first introduce the structure-institutions-agency approach and how it helps make discourse analysis and QCA easier. Then I introduce an example showing how QCA can be used on village data from rural India. The same approach can also be carried out as a national or state-level investigation using triangulation. The third section shows how qualitative interviews can be interrogated using keyness analysis to locate the competing discourses in the society. These are concrete realities. The discourse analysis takes a step-by-step approach and can be used on big data (unstructured data).
My Bangladesh and Indian case material is then examined vis a vis debt outcomes. It contrasts of modernized discourse surrounding NGO memberships. The debates and controversies are part of the interpretive story. In the case of debt, less is said, yet it is a key constraining and enabling factor. An explanatory fuzzy set model is set up to develop an interpretation. These models can focus on evaluation of the differential experience of officially sanctioned debt on households. In rural Bangladesh debts (taken from people or from banks) are a social asset often dominated by men. Yet the access point for some rural credit is a scale of borrowing, which on average exceeds annual rural asset values, and how it overlaps with NGO membership. The project was funded by the ESRC DFID and by the British Academy.
Wendy Olsen is currently writing a book about mixed methods with Sage, and she also wrote ‘Data Collection: Key Debates and Methods in Social Research,’ (London, Sage: 2012), covering several of the fuzzy set and discourse analysis topics in this seminar.