摘要
因果推理有两种基本形式:因果影响和因果机制。因果影响主要是通过变量之间的共变性来确定,大样本的回归分析方法是发现变量间共变性和关系模式的有效工具;而因果机制则是讨论原因变量如何导致结果的过程,小样本的深度案例分析,尤其是过程追踪法是发现和理解因果机制的重要手段。因果影响和因果机制同样重要,大样本的回归分析和小样本的案例研究也各有所长。利用过程追踪法来分析因果机制能够帮助我们认识因果关系的复杂性,有助于理解现实世界中那些重要但稀少的事件。运用过程追踪法应当以理论和变量为指导,尽可能多地发现可观测要素,从而增进因果推理的效度。
Causal effect and causal mechanism are two basic forms of causal inference.While causal effect is mainly defined by the covariation between independent and dependent variables,causal mechanism is to analyze the process that how reasons lead to outcome.Large-N regression method is an effective way to find the covariation and general pattern between variables,whereas small-N in-depth case study,especially process-tracing method,is an important tool to identify causal mechanism.Using process tracing to analyze causal mechanism can benefit our understanding of the complexity of causal relations,and comprehend the significant but rare events in the real world.In order to make sure the inference validity,process tracing should be under the guidance of theory and variables,and contain traceable elements as many as possible.
出处
《世界经济与政治》
CSSCI
北大核心
2010年第4期97-108,共12页
World Economics and Politics
关键词
因果影响
因果机制
过程追踪法
因果推理
causal effect,causal mechanism,process tracing,causal inference