A Critique of Dyadic Design

Dyadic research designs concern data that comprise interactions among actors. They are, without a doubt, the most frequent designs employed in the empirical analysis of international politics. But what do such designs carry with them in terms of theoretical claims and statistical problems? These two issues closely intertwine. When testing hypotheses empirically, the statistical model must be a careful operationalization of the theory under consideration. Given that the theoretical and statistical cannot be separated, we discuss dyadic research designs from these two perspectives. We highlight problems of model misspecification, erroneous assumptions about the independence of events, artificial levels of analysis, and the incoherent treatment of multilateral/multiparty events on the theoretical side. On the statistical side, we stress difficult-to-escape challenges to valid inference.二元研究设计（dyadic research designs）聚焦于包含行为体间互动关系的数据集。毫无疑问，它是国际政治实证分析中最为常用的研究设计。但这类研究设计在理论主张与统计问题层面，会带来哪些相关的议题？这两类问题紧密交织。在实证检验假说时，统计模型必须对所考察的理论进行严谨的操作化定义。鉴于理论维度与统计维度不可分割，本文从上述两个视角对二元研究设计展开探讨。在理论层面，本文着重指出了模型设定偏误、事件独立性假设失当、人为设定的分析层级，以及对多边/多党事件的不一致处理等问题。在统计层面，本文强调了有效推断所面临的难以规避的挑战。