实验设计:一种非模式的方法

EXPERIMENTAL DESIGN: A NON-MODEL APPROACH

实验设计通常被当作一种种设计模式,这种方法可称之为实验设计的模式方法。但本文认为,这种常用的方法事实上存在着不少缺点。基于此,本文尝试提出了一种新的实验设计方法——实验设计的变量方法。实验设计的变量方法的基本思想是,从实验研究的问题出发,根据实验中变量的性质或特点确定实验设计的具体内容。本文最后讨论了实验设计变量方法的优越性。

Experimental design in common textbooks is usually treated as various models. Such a view can be called the Model Approach (MA) of experimental design. Based on a review of the defects of the Model Approach, a new approach which we named Variable Approach ( VA) to experimental design, was proposed in this paper. According to this new approach, variables including independent variables, dependent variables and extraneous variables should be defined first when a researcher designs an experiment. And then contents of experiment design could be determined sequentially in term of their nature or characters. The advantages of the variable approach of experiment design were also discussed at the end of this paper.