Quote of the Week

"There is a persistent division in the social sciences between those who prefer to break their material up into variables and those who prefer dealing with whole cases. In our experience, there are few causes of greater confusion among graduate social scientists, many of whom insist on speaking in the language of variables while working with whole cases, or occasionally vice versa . . . Our view is that there is not one 'right' way to do analysis. Both variable-driven and case-driven research are the products of prior conceptualization and theorization, since neither case nor variable exist as objects. If we are interested in parsimonious explanations and generalization as to what causes what, then it is useful to isolate variables and examine their effects across cases. If we are interested in context and in the complexity of outcomes, then whole cases may yield more insight. So one approach may explain part of the outcome in a large number of cases, while another may explain most of the outcome in a small number of cases."

From Approaches and Methodologies in the Social Sciences edited by Donatella Della Porta and Michael Keating