案例研究概推性的理论逻辑与评价体系——基于公共管理案例研究样本论文的实证分析

案例研究是公共管理研究者运用的一种主要研究方法,但却常常被认为科学性不强。其面临的一项主要指责是案例研究的结果缺乏概推性,即源自一个或少数几个案例研究的结果缺乏普遍意义。本文首先梳理既有案例研究方法论文献对该项挑战的理论回应,分析案例研究结果概推性论述中的三种主要理论逻辑:"自然主义式概推"、"分析概推"和"非正式经验概推"。在借鉴现有理论的基础上,提出一个评价体系,以全面评估案例研究者在处理概推性问题上的严谨性。该体系的三个要素是:研究者"运用案例研究的目的"、"选择案例的理由"及"对案例研究结果适用范围的界定"。在分析上述三要素及其相互关系的基础上,设计出具体的评价指标。然后以发表于五种国际一流公共管理学术期刊上的142篇运用案例研究方法的论文为样本,实证分析公共管理案例研究者在应对概推性难题上的实践。结果表明:在所有样本论文中,八成以上以"一般化"研究为目的,四成以上明确说明了案例选择的理由,三分之一以上采用多案例设计,约四分之一说明了影响案例研究结果适用范围的因素。将2005—2009年样本论文与2000—2004年相比,公共管理案例研究者在处理概推性问题上的严谨程度显著提高,明确说明案例选择理由、采用多案例设计及以"一般化"研究为目的的论文比重大幅上升。最后在讨论实证结果的基础上提出拓展案例研究结果概推性的研究策略。

The Theoretical Logics and Evaluation System of Generalizability of Case Study Research: An Empirical Analysis of Case Study Publications in Public Administration

Case study research is a primary research method employed by public administration researchers,but it is often considered low in scientific quality.A major charge against it is that the results of case study research lack generalizability,namely,the results originating from a single or a few cases lack general significance.This article,first of all,sorts out the theoretical responses to that challenge in the existing literature on case study methodology,and analyzes three major theoretical logics regarding the generalizability of case study results: "naturalistic generalization","analytic generalization",and "informal empirical generalization".Drawing on existing theories,this article proposes a comprehensive evaluation system for assessing the rigor in handling the issue of generalizability by case study researchers.The evaluation system consists of three elements: the purpose of employing case study research by the researcher,the reason for case selection provided by the researcher,and the applicable domain of case study results delineated by the researcher.And the article designs specific indicators based on the analysis of the above three elements and their interrelations.The article then empirically analyzes how public administration researchers deal with the difficult issue of generalizability in practice by examining 142 sample publications using case study method published between 2000 and 2009 in five leading international public administration academic journals.Empirical results show that out of the total sample articles,over eighty percent have "generalizing" research purposes,over forty percent explicitly explain the reasons for case selection,over one third adopt multiple case designs,and about one fourth explain the factors affecting the applicable domain of case study results.When sample articles between 2005 and 2009 are compared with those between 2000 and 2004,it is found that public administration researchers have significantly improved the degree of rigor in handling the issue of generalizability.The proportion of articles rises significantly that explicitly explain the reasons for case selection,adopt multiple case designs,and have "generalizing" research purposes.Finally,this article proposes research strategies for increasing the generalizability of case study results based on discussions of the empirical results.