研究问题与统计方法的有机配合

在实证性研究的设计阶段．研究者应明确研究课题，界定自变量、因变量、控制变量和协变量的操作定义和相互关系，并根据研究课题和变量关系选择合适的统计技术。分别讨论了分析变量关系、探讨组间差异、预测组别隶属以及考察变量结构这四类统计技术解答的研究问题，并通过科研实例介绍了根据研究问题选择统计技术的思路。特别指出。一个研究问题通常可以由多个统计方法加以解决。

MATCH OF RESEARCH QUESTIONS AND STATISTICAL METHODS

In the design stage of an empirical study, researchers are suggested to nail down their research questions and operationalize the independent variables, dependent variables, controlled variables and covariate variables and their relationship. Then choose proper statistical methods according to the research questions and variable relationship. Research questions answered by four categories of statistical methods, i. e. , degree of relationship among variables, significance of group differences, prediction of group membership and structur e, are discussed. The approaches of how to choose statistical methods are introduced with examples in sport sciences. It is stressed that often one research question can be answered by several different approaches of statistical methods.