Abstract: | Event data remains one of the best means for evaluating reciprocity and triangularity in international politics. One difficulty with using this type of data has been its susceptibility to the statistical effects of temporal aggregation. This article examines the concept of the natural interval in event data analysis, specifies how a user-selected aggregation interval affects measured stimulus–response behavior, and proposes a method for calculating a minimum threshold for the natural interval. The article then examines how such a minimum threshold reduces the impact of misspecification on perceived relationships for the Amazon River Basin and the Jordan River Basin. |