心理学期刊中的实证研究论文,很多时候都在检验变量之间的因果关系,但学界对因果研究存在一些不同的看法。本文试图回答下面问题:(1)实验中不能操纵的变量是否可以作为原因?(2)非实验研究能不能检验因果关系?(3)因果分析(尤其是中介分析)是否一定要使用追踪数据?追踪设计的主要目的是什么?通过因果要素和因果推理逻辑的辨析,对前面两个问题都给出了肯定的回答,并讨论了判明变量先后顺序、统计控制无关变量的方法;为了回答问题3,厘清了追踪设计在因果分析中的作用:区分变量的先后顺序、有效获得历时性的影响结果,但即时性的因果影响采用追踪设计可能是不合适的。
Causality is frequently examined in empirical studies in psychology and other social and behavioral sciences. In this paper the following three popular opinions are questioned: (1) A variable is not a cause if it cannot be manipulated in an experiment; (2) A cause and effect conclusion cannot be made from a non-experimental study; and (3) Longitudinal data is necessary for supporting causal inference, especially in a mediation model. The focus is on how researchers can identify causal relations among psychological and social phenomena. According to David Hume, when we say that "X causes Y", we mean that (i) Contiguity: X and Y must be contiguous in time and space; (ii) Succession: X must be prior to Y; and (iii) Constant conjunction: there must be a constant union between X and Y.