选择性偏差问题及其识别

Selection Bias in Comparative Political Studies:What It Is and How to Find It

比较政治学作为政治学科的一个重要分支,其引入和借鉴其他学科研究方法丰富了政治学研究方法的选择,但盲目模仿其他学科"方法驱动型研究创新的成功之道"也影响本学科的研究选题定位和研究结果的信度,同时,忽略不同研究方法内在不足会造成比较政治学学科影响力的下降。社会科学研究中普遍存在的选择性偏差在比较政治学研究中被忽视即为一例,无论是定性研究中的案例选择,还是定量研究中的变量确定,由于研究过程中所选取案例或变量样本对总体不具有代表性,进而导致被低估"关联效应"或伪因果机制,形成比较政治学研究中选择性偏差情形。伴随研究方法的发展,如何有效识别从而规避比较政治学研究中的选择性偏差成为学者关注的焦点。作者基于深入探讨选择性偏差的定义、来源、类型及其影响,并结合案例分析如何在研究中识别选择性偏差存在,旨在澄清学者对选择性偏差认知误区、深化对比较政治学研究方法认知、促进比较政治学不同研究方法的融合,进而推动学科研究进步,亦为比较政治学研究中选择性偏差规避等进一步探索奠定学理基础。

As an important sub-discipline of political science,comparative politics introduces research methods from other subjects for reference.However,if a subject only mimics successful method-driven approaches of others rather than has its own methodology logic,it will bring unintended negative consequences.The selection bias,often ignored in comparative politics in Chinese academia,is a methodological flaw which widely exists in social science research.Since the selection of sample individuals or data cannot represent the population to be analyzed,and thus underestimated correlations and quasi-causal mechanisms are likely to occur.With the development of political research method,the selection bias becomes the focus of researchers' attention.This article first constructs current conceptualizations of selection bias.Thereafter,it systematically analyzes the categories,sources,and influences of selection bias by using specific cases,and pointing out how to find it in the research practice.The article aims to contribute to the new development trend of political methodology that we have experienced during the last years.It also emphasizes that the selection bias in Chinese comparative political studies should be taken into consideration as well as rectification,and then puts forward some detailed suggestions.