试析近年我国社会学定量分析方法应用中的问题

Problems in application of quantitative method to sociological research in China

通过对从1999年初至2001年底,发表于《社会学研究》杂志上的以定量分析方法为主的论文的分析,发现在我国社会学定量分析方法应用中存在着下述问题:不注意统计分析方法应用的前提条件;不能正确地选择和使用描述统计和推论统计的统计指标;用非随机样本的调查结果推论总体;用无效模型进行分析;不能正确地解释计算机统计的结果;不能正确地选择与变量的测量层次相适合的统计分析方法;对计算机统计结果不能进行很好的取舍。这些问题都不同程度地影响了研究结果的正确性。

After a review of papers published in Sociological Research from 1999 to 2001 that used quantitative approach in doing the analysis, following problems were identified in the use of the quantitative approach. Less attention was paid to the premise of applying the approach. Descriptive statistics and inferential statistics were improperly chosen and used. Survey results from non-random samples were used to infer a conclusion about total populations. Invalid statistical modes were used for analysis. Statistical results from computer were incorrectly interpreted. Statistical methods were not appropriate for measuring the dependent variable(s) in research question. Results from the computer analysis were not selected properly. All these problems affect the validity of the research results to certain degree.