Varieties of description in political science

Markus Kreuzer writes:

I am organizing a panel at next year’s American Political Science Association meeting tentatively entitled “Varieties of Description.” The idea is to compare and contrast the ways in which different disciplines approach descriptive inferences, that how they go about collective data, how they validate descriptive inferences and what ontological assumptions they make. The panel tries to expand on the work by John Gerring and others that identify description as distinct elements to social analysis. The panel currently has contributors who analyze the structure of historical description as well as thick anthropological-like description. To round out the panel, I am interested in finding contributors who take a more quantitative approach to description. This could involve the use descriptive statistics or exploratory data analysis to describe how, when and where social phenomena unfold or who employ it to generate new theoretical insights. Or it could involve work on concept formation or typologizing.

I responded by pointing him to my paper with Basbøll in which we say that that storytelling, like exploratory data analysis more generally, can be seen as a form of model checking: an interesting story is one which acts as a counterexample to some model (explicit or implicit) of interest.

As we write in our article:

One might imagine a statistician criticizing storytellers for selection bias, for choosing the amusing, unexpected, and atypical rather than the run-of-the- mill boring reality that should form the basis for most of our social science. But then how can we also say the opposite that stories benefit from being anomalous? We reconcile this apparent contradiction by placing stories in a different class of evidence from anecdotal data as usually conceived. The purpose of a story is not to pile on evidence in support of one theory or another but rather to shine a spotlight on an anomaly—a problem with an existing model—and to stand as an immutable object that conveys the complexity of reality.

Anyway, if anyone is interested in participating in Kreuzer’s panel at next year’s APSA meeting, feel free to contact him directly.

The post Varieties of description in political science appeared first on Statistical Modeling, Causal Inference, and Social Science.