QUANTITATIVE ETHNOGRAPHY

An academic study explains how ethnography’s cultural analysis can deepen the mining of big data. 

In the last 20 years, there have been two parallel revolutions ripe for dovetailing: the movement among the sciences toward the analysis of increasingly complex systems demanding disciplinary collaboration and the birth of big data, the expert collection and mining of infinite piles of statistical minutiae. According to the author, the difficulty with data analysis today is that despite its increasing sophistication, it still needs help managing the “nasty problem” of human complexity—the fact that all of this statistical fodder requires organization and interpretation in order to be transformed into real understanding. Ethnography has a pedigree of doing precisely this: It not only carefully draws generalizations about human behavior from specific data (this makes it “microgenetic”), but it also plumbs cultural questions. Shaffer (How Computer Games Help Children Learn, 2006) proposes a way the techniques of ethnography, which produces a “thick,” qualitative analysis rich with cultural significance, can be combined with the “thin,” quantitative analysis of big data: “To do anything less—to pretend that the mountains of data are not islands in a sea of cultural significance—may be mathematically rigorous, but in the end is conceptually empty.” The author, an ethnographer trained at the prestigious Media Lab at the Massachusetts Institute of Technology, provides an impressively comprehensive account of what ethnography is. He explores the way the discipline handles the inevitable problem of bias, the use of statistical models, the nature of generalization, and the basic methods of analysis. Shaffer also examines the theoretical nature of the connection between qualitative and quantitative analysis and the potential pitfalls involved. His study is clearly designed for other academics—average readers won’t be gripped by lengthy discussions about epistemic frameworks—but his prose is startlingly accessible and is likely to appeal to anyone with an amateur interest in the systematic examination of culture. This is a timely and original work and should be required reading for ethnographers and statisticians alike. 

A valuable contribution to the debate regarding the future of big data.