摘要
以往在对特定的方差分析（ANOVA）资料指定其数学模型时，都是根据因子及因子之间的互作效应的多少来进行的，似乎在ANOVA问题当中不存在模型选择问题。本文则表明在ANOVA当中不但存在着模型选择问题，而且利用Cp准则还能导致正确的结论。
In this article, An important procedure Cp Criterion with which to select best linear models is introduced in detail, and then it was used to opt for best ANOVA model. Results demonstrate that the application of CP-statistic to model selection can lead to correct conclusions
出处
《内蒙古民族大学学报》
1997年第4期7-10,共4页
Journal of Inner Mongolia University for the Nationalities
