资源描述
试验设计作业
1、下表为小麦栽培试验的产量结果(kg),随机区组设计,小区计产面积为12m2,试作分析。在表示最后结果时需化为每亩产量(kg)。假定该试验为一完全随机设计,试分析后将其试验误差与随机区组时的误差作一比较,看看划分区组的效果如何?
处 理
区 组
Ⅰ
Ⅱ
Ⅲ
Ⅳ
A
6.2
6.6
6.9
6.1
B
5.8
6.7
6.0
6.3
C
7.2
6.6
6.8
7.0
D
5.6
5.8
5.4
6.0
E
6.9
7.2
7.0
7.4
F
7.5
7.8
7.3
7.6
完全随机设计的程序如下:
data li_1;
do i=1 to 6;
do j=1 to 4;
input x@@;
output;
end;
end;
cards;
6.2 6.6 6.9 6.1
5.8 6.7 6 6.3
7.2 6.6 6.8 7
5.6 5.8 5.4 6
6.9 7.2 7 7.4
7.5 7.8 7.3 7.6
;
proc anova;
class i;
model x=i;
means i;
run;
SAS输出结果如下: Sum of
Source DF Squares Mean Square F Value Pr > F
Model 5 8.97208333 1.79441667 20.87 <.0001
Error 18 1.54750000 0.08597222
Corrected Total 23 10.51958333
R-Square Coeff Var Root MSE x Mean
0.852893 4.406415 0.293210 6.654167
Source DF Anova SS Mean Square F Value Pr > F
i 5 8.97208333 1.79441667 20.87 <.0001
随机区组设计的程序如下:
data li_3;
do i=1 to 6;
do j=1 to 4;
input x@@;
output;
end;
end;
cards;
6.2 6.6 6.9 6.1
5.8 6.7 6 6.3
7.2 6.6 6.8 7
5.6 5.8 5.4 6
6.9 7.2 7 7.4
7.5 7.8 7.3 7.6
;
proc anova;
class i j;
model x=i j;
run;
结果如下:
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 8 9.24333333 1.15541667 13.58 <.0001
Error 15 1.27625000 0.08508333
Corrected Total 23 10.51958333
R-Square Coeff Var Root MSE x Mean
0.878679 4.383576 0.291690 6.654167
Source DF Anova SS Mean Square F Value Pr > F
i 5 8.97208333 1.79441667 21.09 <.0001
j 3 0.27125000 0.09041667 1.06 0.3943
结果分析:随机区组设计的误差要小一些。
2、下表为水稻品种比较试验的产量结果(kg),5×5拉丁方设计,小区计产面积30m2,试分析。
B
25
E
23
A
27
C
28
D
20
D
22
A
28
E
20
B
28
C
26
E
18
B
25
C
28
D
24
A
25
A
26
C
26
D
22
E
19
B
24
C
23
D
23
B
26
A
33
E
20
拉丁方设计的程序如下:
data li_4;
do i=1 to 5;
do j=1 to 5;
input fz $ x@@;
output;
end;
end;
cards;
B 25 E 23 A 27 C 28 D 20
D 22 A 28 E 20 B 28 C 26
E 18 B 25 C 28 D 24 A 25
A 26 C 26 D 22 E 19 B 24
C 23 D 23 B 26 A 33 E 20
;
proc anova;
class i j fz;
model x=i j fz;
run;
结果如下:
The ANOVA Procedure
Dependent Variable: x
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 12 255.2800000 21.2733333 7.40 0.0008
Error 12 34.4800000 2.8733333
Corrected Total 24 289.7600000
R-Square Coeff Var Root MSE x Mean
0.881005 6.958501 1.695091 24.36000
Source DF Anova SS Mean Square F Value Pr > F
i 4 8.5600000 2.1400000 0.74 0.5798
j 4 44.5600000 11.1400000 3.88 0.0302
fz 4 202.1600000 50.5400000 17.59 <.0001
结果分析:列间,品种间差异是显著的。
3、左下表为玉米播期试验结果,右下表为油菜品比试验结果,皆为随机区组设计,试试分析。
播期
区组
Ⅰ
Ⅱ
Ⅲ
A
20.3
22.1
20.7
B
19.8
19.0
18.6
C
18.4
16.8
17.4
D
16.0
16.6
18.1
E
15.2
14.9
15.3
F
14.9
15.9
14.0
G
14.0
18.8
15.3
品 种
区
组
Ⅰ
Ⅱ
Ⅲ
Ⅳ
A
3.9
3.9
3.6
4.5
B
5.8
6.3
5.9
6.8
C
4.4
4.4
5.6
4.5
D
5.5
5.2
5.4
6.7
E
6.8
6.9
7.4
6.0
F
7.3
7.2
7.5
7.0
玉米播期试验的产量(kg)结果 油菜品比试验的产量(kg)结果
玉米的随机区组设计程序如下:
data li_3;
do i=1 to 7;
do j=1 to 3;
input x@@;
output;
end;
end;
cards;
20.3 22.1 20.7
19.8 19 18.6
18.4 16.8 17.4
16 16.6 18.1
15.2 14.9 15.3
14.9 15.9 14
14 18.8 15.3
;
proc anova;
class i j;
model x=i j;
run;
结果如下:
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 8 90.6942857 11.3367857 7.61 0.0011
Error 12 17.8771429 1.4897619
Corrected Total 20 108.5714286
R-Square Coeff Var Root MSE x Mean
0.835342 7.078630 1.220558 17.24286
Source DF Anova SS Mean Square F Value Pr > F
i 6 88.17142857 14.69523810 9.86 0.0005
j 2 2.52285714 1.26142857 0.85 0.4529
油菜随机区组设计程序如下:
data li_3;
do i=1 to 6;
do j=1 to 4;
input x@@;
output;
end;
end;
cards;
3.9 3.9 3.6 4.5
5.8 6.3 5.9 6.8
4.4 4.4 5.6 4.5
5.5 5.2 5.4 6.7
6.8 6.9 7.4 6
7.3 7.2 7.5 7
;
proc anova;
class i j;
model x=i j;
run;
结果如下:
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 8 31.27500000 3.90937500 14.18 <.0001
Error 15 4.13458333 0.27563889
Corrected Total 23 35.40958333
R-Square Coeff Var Root MSE x Mean
0.883235 9.097702 0.525013 5.770833
Source DF Anova SS Mean Square F Value Pr > F
i 5 30.81708333 6.16341667 22.36 <.0001
j 3 0.45791667 0.15263889 0.55 0.6535
结果分析:两播种期试验均表明,不同播种期对试验结果的影响是显著的。
4、下表为水稻栽培试验的小区产量(kg)结果,5×5拉丁方设计,,试分析。。
B 14
E 15
C 25
A 12
D 16
E 18
D 12
B 15
C 14
A 11
C 21
A 13
D 13
B 13
E 19
A 10
C 24
E 18
D 14
B 12
D 12
B 15
A 11
E 20
C 26
拉丁方设计程序如下:
data li_4;
do i=1 to 5;
do j=1 to 5;
input fz $ x@@;
output;
end;
end;
cards;
B 14 E 15 C 25 A 12 D 16
E 18 D 12 B 15 C 14 A 11
C 21 A 13 D 13 B 13 E 19
A 10 C 24 E 18 D 14 B 12
D 12 B 15 A 11 E 20 C 26
;
proc anova;
class i j fz;
model x=i j fz;
run;
结果如下:
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 12 401.9200000 33.4933333 4.41 0.0078
Error 12 91.1200000 7.5933333
Corrected Total 24 493.0400000
R-Square Coeff Var Root MSE x Mean
0.815187 17.52926 2.755600 15.72000
Source DF Anova SS Mean Square F Value Pr > F
i 4 23.0400000 5.7600000 0.76 0.5717
j 4 17.0400000 4.2600000 0.56 0.6954
fz 4 361.8400000 90.4600000 11.91 0.0004
结果分析:品种间差异对结果的影响是显著的。
5、 调查某队元麦及元麦和蚕豆混种、间种田块的产量(混、间种者为麦、豆产量合计),得结果(kg/0.1亩)于表:(1)设以元麦单种为对照,试以LSD法作多重比较;(2)设预定要作的比较是单种对混、间种,混种对间种,2麦1豆间种对3麦2豆间种,试作单一自由度的独立比较。
元麦单种
麦豆混种
2麦1豆间种
3麦2豆间种
20
24
30
30
24
23
28
33
22
28
34
31
18
21
32
36
21
24
31
35
先作方差分析:
data li_1;
do i=1 to 4;
do j=1 to 5;
input x@@;
output;
end;
end;
cards;
20 24 22 18 21
24 23 28 21 24
30 28 34 32 31
30 33 31 36 35
;
proc anova;
class i;
model x=i;
means i;
run;
结果如下:
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 3 483.7500000 161.2500000 28.04 <.0001
Error 16 92.0000000 5.7500000
Corrected Total 19 575.7500000
R-Square Coeff Var Root MSE x Mean
0.840208 8.799691 2.397916 27.25000
Source DF Anova SS Mean Square F Value Pr > F
i 3 483.7500000 161.2500000 28.04 <.0001
The ANOVA Procedure
Level of --------------x--------------
i N Mean Std Dev
1 5 21.0000000 2.23606798
2 5 24.0000000 2.54950976
3 5 31.0000000 2.23606798
4 5 33.0000000 2.54950976
处理
均数
差数
均数-21
均数-24
均数-31
3麦2豆间种
33
12(极显著)
9(极显著)
2
2麦1豆间种
31
10(极显著)
7(极显著)
麦豆混种
24
3
元麦单种
21
6. 有一大豆试验,A因素为品种,有A1、A2、A3、A4 4个水平,B因素为播期,有B1、B2、B3 3个水平,随机区组设计,重复3次,小区计产面积25平方米,其田间排列和产量(kg)如下图,试作方差分析。检验:品种、播期,品种×播期的效应是否显著?
区组Ⅰ
A1B1
12
A2B2
13
A3B3
14
A4B2
15
A2B1
13
A4B3
16
A3B2
14
A1B3
13
A4B1
16
A1B2
12
A3B1
14
A2B3
14
区组Ⅱ
A4B2
16
A1B3
14
A2B1
14
A3B3
15
A1B2
12
A2B3
13
A4B1
16
A3B2
13
A2B2
13
A3B1
15
A1B1
13
A4B3
17
区组Ⅲ
A2B3
13
A3B1
15
A1B2
11
A2B1
14
A4B3
17
A3B2
14
A2B2
12
A4B1
15
A3B3
15
A1B3
13
A4B2
15
A1B1
13
DATA CaP;
DO a=1 to 4;
DO b=1 to 3;
DO n=1 to 3;
input y@@;
output;
end;
end;
end;
DROP n;
CARDS;
12 16 13
13 14 15
14 14 11
15 15 14
13 12 17
16 13 14
14 16 12
13 13 15
16 13 15
12 15 13
14 13 15
14 17 13
;
PROC ANOVA; CLASS A B; MODEL y=A B A*B;
RUN;
结果如下:
The ANOVA Procedure
Dependent Variable: y
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 11 9.33333333 0.84848485 0.31 0.9778
Error 24 66.66666667 2.77777778
Corrected Total 35 76.00000000
R-Square Coeff Var Root MSE y Mean
0.122807 11.90476 1.666667 14.00000
Source DF Anova SS Mean Square F Value Pr > F
a 3 2.88888889 0.96296296 0.35 0.7918
b 2 0.50000000 0.25000000 0.09 0.9142
a*b 6 5.94444444 0.99074074 0.36 0.8989
结果分析:品种、播期,品种×播期对试验的影响均不显著。
7. 有一小麦裂区试验,主区因素A,分A1(深耕)、A2(浅)两水平,副区因素B,分B1(多肥)、B2(少肥)两水平,重复3次,小区计产面积15平方米,其田间排列和产量(假设数字)如下图,试作方差分析。因素A, 因素B及其交互效应是否显著?
A1
A2
A2
A1
A2
A1
B1
9
B1
7
B2
3
B1
11
B2
1
B2
4
B2
6
B2
2
B1
5
B2
4
B1
6
B1
12
区组Ⅰ
区组Ⅱ
区组Ⅲ
解:程序如下:
title '裂区试验的统计分析';
data lq;
do a=1 to 2;
do b=1 to 2;
do g=1 to 3;
input yield @; output;
end;
end;
end;
cards;
9 11 12
6 4 4
7 5 6
2 3 1;
proc anova;
class g a b;
model yield= g a b ;
means a b/duncan;
run;
proc anova;
class g a b;
model yield= g a b a*b g*a g*a*b;
test H=g a E=g*a;
test H=b a*b E=g*a*b;
run;
试验结果:
裂区试验的统计分析 2 08:07 Wednesday, July 3, 2002
Analysis of Variance Procedure
Dependent Variable: YIELD
Source DF Sum of Squares F Value Pr > F
Model 4 115.50000000 14.27 0.0018
Error 7 14.16666667
Corrected Total 11 129.66666667
R-Square C.V. YIELD Mean
0.890746 24.38754 5.83333333
Source DF Anova SS F Value Pr > F
G 2 0.16666667 0.04 0.9599
A 1 40.33333333 19.93 0.0029
B 1 75.00000000 37.06 0.0005
裂区试验的统计分析 6 08:07 Wednesday, July 3, 2002
Analysis of Variance Procedure
Dependent Variable: YIELD
Source DF Sum of Squares F Value Pr > F
Model 11 129.66666667 . .
Error 0 .
Corrected Total 11 129.66666667
R-Square C.V. YIELD Mean
1.000000 0 5.83333333
Source DF Anova SS F Value Pr > F
G 2 0.16666667 . .
A 1 40.33333333 . .
B 1 75.00000000 . .
A*B 1 3.00000000 . .
G*A 2 1.16666667 . .
G*A*B 4 10.00000000 . .
Tests of Hypotheses using the Anova MS for G*A as an error term
Source DF Anova SS F Value Pr > F
G 2 0.16666667 0.14 0.8750
A 1 40.33333333 69.14 0.0142
Tests of Hypotheses using the Anova MS for G*A*B as an error term
Source DF Anova SS F Value Pr > F
B 1 75.00000000 30.00 0.0054
A*B 1 3.00000000 1.20 0.3349
结果分析:因素A、因素B的显著性水平为0.0029、0.0005,是显著的,而AXB的F值是1.2,顾AXB是不显著的。B1与B2有显著的差别,A1与A2有显著的差别,最好的是A1B1的组合。
8. 设若上题小麦耕深与施肥量试验为条区设计,田间排列和产量将相应如下图,试作分析,并与裂区设计结果相比较)。
A1
A2
A2
A1
A2
A1
B1
9
7
B1
5
11
B2
1
4
B2
6
2
B2
3
4
B1
6
12
解:程序如下:
title '条区试验的统计分析';
data tq;
do g=1 to 3;
do a=1 to 2;
do b=1 to 2;
input yield @; output;
end;
end;
end;
cards;
9 6 7 2
11 4 5 3
12 4 6 1;
proc anova;
class g a b;
model yield= g a b ;
means a b/duncan;
run;
展开阅读全文