《用spss20进行可重复单因素随机区组、两因素随机区组、两因素裂区试验设计的方差分析.doc》由会员分享，可在线阅读，更多相关《用spss20进行可重复单因素随机区组、两因素随机区组、两因素裂区试验设计的方差分析.doc（14页珍藏版）》请在金锄头文库上搜索。
1、一、可重复单因素随机区组试验设计8个小麦品种的产比试验，采用随机区组设计，3次重复，计产面积25平米，产量结果如下，进行方差分析和多重比较。表1 小麦品比试验产量结果（公斤）品种区组产量1110.92110.83111.1419.15111.86110.17110819.3129.12212.33212.54210.75213.96210.67211.58210.41312.223143310.54310.15316.86311.87314.18314.41、打开程序把上述数据输入进去。2、执行：分析-一般线性模型-单变量。3、将产量放进因变量，品种和区组放进固定因子。4、单击模型，选择设定单
2、选框，将品种和区组放进模型中，只分析主效应。5、在两两比较中进行多重比较，这里只用分析品种。可以选择多种比较方法。6、分析结果。主体间效应的检验因变量: 产量源III 型平方和df均方FSig.校正模型61.641a96.8494.174.009截距3220.16713220.1671962.448.000区组27.561213.7808.398.004品种34.08074.8692.967.040误差22.972141.641总计3304.78024校正的总计84.61323a. R 方 = .729（调整 R 方 = .554）这里只须看区组和品种两行，两者均达到显著水平，说明土壤肥力和品
3、种均影响产量结果。下面是多重比较，只有方差分析达到显著差异才进行多重比较。多个比较因变量: 产量(I) 品种(J) 品种均值差值 (I-J)标准 误差Sig.95% 置信区间下限上限LSD1.002.00-1.63331.04591.141-3.8766.60993.00-.63331.04591.555-2.87661.60994.00.76671.04591.476-1.47663.00995.00-3.4333*1.04591.005-5.6766-1.19016.00-.10001.04591.925-2.34332.14337.00-1.13331.04591.297-3.37661
4、.10998.00-.63331.04591.555-2.87661.60992.001.001.63331.04591.141-.60993.87663.001.00001.04591.355-1.24333.24334.002.4000*1.04591.038.15674.64335.00-1.80001.04591.107-4.0433.44336.001.53331.04591.165-.70993.77667.00.50001.04591.640-1.74332.74338.001.00001.04591.355-1.24333.24333.001.00.63331.04591.55
5、5-1.60992.87662.00-1.00001.04591.355-3.24331.24334.001.40001.04591.202-.84333.64335.00-2.8000*1.04591.018-5.0433-.55676.00.53331.04591.618-1.70992.77667.00-.50001.04591.640-2.74331.74338.00.00001.045911.000-2.24332.24334.001.00-.76671.04591.476-3.00991.47662.00-2.4000*1.04591.038-4.6433-.15673.00-1.
6、40001.04591.202-3.6433.84335.00-4.2000*1.04591.001-6.4433-1.95676.00-.86671.04591.421-3.10991.37667.00-1.90001.04591.091-4.1433.34338.00-1.40001.04591.202-3.6433.84335.001.003.4333*1.04591.0051.19015.67662.001.80001.04591.107-.44334.04333.002.8000*1.04591.018.55675.04334.004.2000*1.04591.0011.95676.
7、44336.003.3333*1.04591.0071.09015.57667.002.3000*1.04591.045.05674.54338.002.8000*1.04591.018.55675.04336.001.00.10001.04591.925-2.14332.34332.00-1.53331.04591.165-3.7766.70993.00-.53331.04591.618-2.77661.70994.00.86671.04591.421-1.37663.10995.00-3.3333*1.04591.007-5.5766-1.09017.00-1.03331.04591.34
8、0-3.27661.20998.00-.53331.04591.618-2.77661.70997.001.001.13331.04591.297-1.10993.37662.00-.50001.04591.640-2.74331.74333.00.50001.04591.640-1.74332.74334.001.90001.04591.091-.34334.14335.00-2.3000*1.04591.045-4.5433-.05676.001.03331.04591.340-1.20993.27668.00.50001.04591.640-1.74332.74338.001.00.63
9、331.04591.555-1.60992.87662.00-1.00001.04591.355-3.24331.24333.00.00001.045911.000-2.24332.24334.001.40001.04591.202-.84333.64335.00-2.8000*1.04591.018-5.0433-.55676.00.53331.04591.618-1.70992.77667.00-.50001.04591.640-2.74331.7433基于观测到的均值。 误差项为均值方 (错误) = 1.641。*. 均值差值在 0.05 级别上较显著。产量品种N子集12Duncana,
10、b4.0039.96671.00310.73336.00310.83333.00311.36678.00311.36677.00311.866711.86672.00312.366712.36675.00314.1667Sig.060.055已显示同类子集中的组均值。 基于观测到的均值。 误差项为均值方 (错误) = 1.641。a. 使用调和均值样本大小 = 3.000。b. Alpha = 0.05。二、两因素可重复随机区组试验设计下面是水稻品种和密度对产量的影响，采用随机区组试验设计，3次重复，品种3个水平，密度3个水平，共27个观测值。小区计产面积20平米。表2 水稻品种与密度产比试验
11、品种密度区组产量11181128113812171227123613161325133621192129213822172229223623182327233631173127313632183227323833110332933391、输入数据，执行：分析-一般线性模型-单变量。注意区组作为随机因子。2、选择模型。注意模型中有三者的主效和品种与密度的交互。 3、分析结果。注意自由度的分解。使用一个误差（0.486）计算F值。主体间效应的检验因变量: 产量源III 型平方和df均方FSig.截距假设1496.33311496.3331035.923.001误差2.88921.444a品种假设6.22223.1116.400.009误差7.77816.486b密
