从一个到全体－质的研究结果的推论问题

From One to All：the Generalizability of Qualitative Research

在量的研究中，[推论]是一个不言自明的概念，指的是用概率抽样的方法抽取一定的样本量以后，将所获得的研究结果推论到从中抽样的总体。[推论]是量的研究一个必不可少的前提，强调的是研究结果的普适性。质的研究选择的样本一般都比较小，有时甚至只有一个人或一个地点，而且抽样时遵循的是[目的性抽样]的原则，即根据具体情境抽取可以为研究问题提供最大信息量的人和事(Patton，1990：169--186)。

The issue of generalizability has been a tough topic in qualitative research. This paper take a fresh look at this issue. The author argues that generalizahility in qualitative research is not so much on verifying things already known to the public， as on finding something new so as to expand the knowledge structure of mankind. The endeavor is undertaken in a ＂naturalistic＂ process， in which the qualitative researcher tries to empathize with the researched， explore his/her own reflexivity in the interaction， and provide a ＂thick description＂ of the findings located in the actual contexts.