本文试图说明 ,从作为大多数中国内地问卷调查研究之基础的对中国人口的非代表性抽样能够 (及不能够 )推出什么结论。显然 ,出于方便而选择的一个或几个地区内的被访问的中国人样本 ,不能可靠地代表研究者感兴趣的更大的总体。本文指出 ,原则上没有理由认为地区样本是“个案研究”。尽管我们不能依靠来自地区样本的数据沿着任何单一的维度去描述更大的总体 ,这种非代表性并不一定影响有关变量之间关系的研究结果的可推广性。 This article attempts to show what conclusions can (and can not) be drawn from the non-representative sampling of the Chinese population that forms the basis of most of the mainland Chinese questionnaire studies. Obviously, the sample of Chinese people interviewed in one or a few regions chosen for convenience can not reliably represent the larger totality that the researcher is interested in. This article points out that in principle there is no reason to think of a regional sample as a “case study.” Although we can not rely on data from regional samples to describe larger totals along any single dimension, this unrepresentativeness does not necessarily affect the generalizability of the findings on the relationship between variables.