Conflict datasets: Definitions and measurement

Publisher Website
Google Scholar
Abstract
Data‐making in international politics is—explicitly or implicitly—driven by theory. A certain set of premises defines for the investigator what needs to be observed, how to observe, and how much to observe. Accordingly, datasets on international conflict differ from one another quite sharply along several dimensions. The consumer's choice of an appropriate dataset must be guided not only by the practical question of which dataset is available, but also by the question of which dataset is the most suitable in terms of the nature of its underlying theoretical assumptions. The paper discusses (1) the characteristics of some, but by no means all, conflict datasets and their underlying theoretical premises; (2) how these premises affect the results of the data‐generation effort and how differences in underlying premises result in substantial cross‐data differences; and (3) some possible directions toward standardizing conflict data in international politics as well as some inherent difficulties.

This publication has 12 references indexed in Scilit: