摘要
比较案例研究是社会科学方法中最为核心的问题之一,密尔方法的应用是比较案例研究走向科学化的重要标志。当前学界对于比较案例研究中密尔方法的认知仍存在困惑,即基于小样本的密尔方法何以得出比大样本的统计分析更为确定性的结论,因而,有必要澄清比较案例研究的或然性。无论是案例研究还是定量分析,都是或然性的,并且这两种传承的基础逻辑是一致的。比较案例研究的或然性源于社会科学与自然科学的差异,使得研究设计的各个步骤都产生确定性损失。具体而言,这种确定性损失来自于内部效度和外部效度:前者包括了对事实认定、寻找和检验机制、过程追踪与强化信念;后者包括了变量赋值、基于布尔代数的因果推断和理论适用范围的推广,等等。因而,研究者在进行比较案例研究时,更加适合使用贝叶斯逻辑,将因果过程观察值和数据集观察值视为调整对理论主观信心的证据,而非在小样本的基础上去寻求确定性的解释。
Comparative case studies(CCS)is one of the most important issues in social science methodology.However,scholars are confused about why the Mills method based on small samples can come to a more deterministic conclusion than the statistical analysis of large samples,and thus it is necessary to clarify the probability of CCS.Both case studies and quantitative analysis are probable,and the basic logic of the two cultures is the same.The probability of CCS stems from the differences between social sciences and natural sciences,which lead to deterministic losses in all steps of research design.The deterministic loss results from internal validity and external validity:the former includes fact finding,testing mechanism and process tracing;the latter includes variable assignment,causal inference based on Boolean algebra and the extension of the application of the theory.Consequently,CCS are more applicable to Bayesian logic,viewing causal process observations and dataset observations as evidence for adjusting subjective confidence in the theory,rather than seeking deterministic explanations on the basis of small samples.
作者
叶成城
曹航
Ye Chengcheng;Cao Hang(Institute of International Studies,Shanghai Academy of Social Sciences,Shanghai 200020;Institute of the Belt and Road Studies,Yunnan University,Kunming 650091,China)
出处
《复旦学报（社会科学版）》
北大核心
2023年第6期171-179,共9页
Fudan Journal（Social Sciences）
基金
国家社科基金后期资助项目“早期西欧国家的构建与兴衰(1492-1848)”(项目批准号:21FGJB006)的阶段性成果。
