论文部分内容阅读
同一年在不同生态地区进行相同的试验(如作物品种比较试验或区域试验,某项农业措施对作物的增产增益效应等),均应进一步对多点试验的结果作联合分析,以测验品种(或处理)间的差异是否显著,测验品种(或处理)与地区的交互作用(即品种或处理是否因试验地区的不同而异)。此法,是在各试点分别作方差分析的基础上,首先应进一步对各试点的误差均方作同质性检验(齐性检验),若同质,便可将备试点的结果进行综合分析;如不同质,就应考虑剔除个别试验误差偏大的试点,使误差均方同质后再进行综合分析。本文例举的综合分析方法,以处理为固定模型,地区和区组为随机模型。若测验结果其处理均方显著地高于处理与地区
In the same year in different ecoregions of the same test (such as comparison of crop varieties or regional trials, the effect of an agricultural measure on crop yield enhancement, etc.) should be further multi-point test results for joint analysis to test varieties ( Or treatment) is significant, and the test breed (or treatment) interacts with the region (ie, whether the breed or treatment varies by region of test). This method is based on the variance analysis of each pilot on the basis of the first of all should be further tested for homogeneity of the pilot (homogeneity test), if homogeneous, the results of the pilot can be a comprehensive analysis ; If different quality, it should be considered excluding individual test error is too large pilot, so that the error is homogeneous after a comprehensive analysis. This article presents a comprehensive analysis of methods to deal with stochastic models for fixed models, regions and blocks. If the test results were significantly higher than the mean square of treatment and area