---

作 者 ：王金明

期 刊 ：作物学报 1992年 18卷 04期 页码：291-295

关键词：田间试验分段设计；

Keywords:Segmented experiment design,

摘 要 ：为了在两个方向上控制田间试验的地力差异,而又不过分限制处理数m的取值,本文提出了一种变形拉丁方设计的试验方法。该种设计是在随机区组设计的基础上,将n个区组做为行区组看待。再将这n个行区组的m个小区分成n个段作为列区组来布置试验,并使所有处理在每个行区组和每个列区组中都只出现一次。在方差分析中,将行区组和列区组

Abstract:In order to control the difference of soil fertility in two directions and at the sametime to have enough number of treatments, a variant of Latin square design was develo-ped.This method was designed on the basis of randomized block design. The nblocks were considered as row blocks. Then the n bloc

---