Abstract:
Wen et al. (2024), in their paper published in Acta Psychologica Sinica, focused on the long-term ambiguous use of the word “influence” and proposed the concept of “influence relationship,” which may change the pragmatic practice of Chinese psychology in the future. However, there are some doubts in that paper: (A) The word “influence” is regarded as causal language in both Chinese and English, in both public and academic circles; thus, it is difficult to be persuaded by the authors that it does not convey causal meaning. (B) A clear definition of “influence relationship” was absent in that paper. This concept seems to have no essential difference from causal relationship. (C) That paper confused the goal with the means to achieve the goal. It is weird to create a third kind of goal that is different from causation and correlation just because the means of some studies cannot provide causal evidence for the causal goal. More importantly, the authors repeatedly iterated that the covariant and directional variable relationship was not named; however, this relationship has been called “prediction” by the academic community. Accordingly, this commentary is intended to remind researchers to consider other possible ways of describing relevant situations, and to make prudent decisions about whether to introduce so-called “influence relationships” into their research.
Expert Opinion:本稿件对最近发表在《学报》上可能产生重大影响的《变量之间的影响关系和多重影响因素的共同作用类型》进行了回应,本人认为及时且必要。稿件本身对于提升国内健康的学术争鸣具有重要意义!以下为本人通读本稿件以及所对应的《变量之间的影响关系和多重影响因素的共同作用类型》后的观点,供作者与编辑部参考:
第一、本稿件的三点质疑均合理,尤其“该文混淆了目标与实现目标的手段”这一点在科学心理学中尤其关键,因此本人认为应该将这一点是最重要的质疑,可以考虑提到稿件中更突出的位置。
第二,接上一点的“手段”,事实上,其他领域在手段上已经有了不少进展,心理学研究者不能对此视而不见。在统计学中,因果推断在近年来有不少新的进展,有一些相关的科普图书,例如计算机的图灵奖获得者Pearl的the book of why。不少国际心理学同行已经在积极地将这些统计上的因果推断框架引入到心理学和社会科学研究中(如本稿件中已经引用的Rohrer, 2018,也包括Rohrer后续的工作,包括2024年的《Causal inference for psychologists who think that causal inference is not for them》(https://onlinelibrary.wiley.com/doi/abs/10.1111/spc3.12948),也有国内同行在关注临床中的因果推断,如(Wang, Q., Wang, Q., & Zhang, R.-Y. (2023). Claim causality with clarity. Psychoradiology, 3, kkad007. https://doi.org/10.1093/psyrad/kkad007)。在此背景下,如果心理学研究者仍采用暧昧不清的语言去回避因果推断,可能会阻碍心理学的发展。
第三,因果关系的定义中和建议的行为中,稿件中可考虑增加更多其他领域关于因果推断的探讨,鼓励学生或者青年学者超越心理学的范围,广泛关注研究方法的进展。例如,Pearl等关于因果推断的内容,包括Structural Causal Model (SCM),如(Pearl, 2009, doi: https://doi.org/10.1214/09-SS057; Ding, 2024, url: https://www.amazon.com/Course-Inference-Chapman-Statistical-Science/dp/1032758627)。