Está en la página 1de 38

Cuadrado latino

options ls=120 ps=1000;


data cl;
input turno tec met$ tiempo;
cards;
1
1
3
90
1
2
4
96
1
3
1
84
1
4
2
88
2
1
2
90
2
2
3
91
2
3
4
96
2
4
1
88
3
1
1
89
3
2
2
97
3
3
3
98
3
4
4
98
4
1
4
104
4
2
1
100
4
3
2
104
4
4
3
106
;
proc glm;
class turno tec met;
model tiempo=turno tec met/ss3;
random turno tec/test;
lsmeans met/tdiff;
contrast '1-2' met -1 1 0 0;
estimate '1-2' met -1 1 0 0;
run;

The SAS System

09:32 Wednesday, October 20, 2014 1


The GLM Procedure
Class Level Information
Class
turno
tec
met

Levels
4
4
4

Values
1234

1234
1234

Number of observations

16

The SAS System

09:32 Wednesday, October 20, 2014 2

The GLM Procedure


Dependent Variable: tiempo
Source

DF

Sum of
Squares

Model

630.0625000

Error

Corrected Total

15

R-Square

turno
tec
met

3
3
3

18.36

Pr > F
0.0011

3.8125000

Root MSE

2.056682
DF

70.0069444

F Value

652.9375000

Coeff Var

0.964966
Source

22.8750000

Mean Square

1.952562

Type III SS
467.1875000
17.1875000
145.6875000

tiempo Mean
94.93750

Mean Square

F Value

Pr > F

155.7291667
40.85 0.0002
5.7291667
1.50 0.3065
48.5625000
12.74 0.0052

The SAS System

09:32 Wednesday, October 20, 2014 3

The GLM Procedure


Source
turno

Type III Expected Mean Square


Var(Error) + 4 Var(turno)

tec

Var(Error) + 4 Var(tec)

met

Var(Error) + Q(met)

The SAS System

09:32 Wednesday, October 20, 2014 4

The GLM Procedure


Tests of Hypotheses for Mixed Model Analysis of Variance
Dependent Variable: tiempo
Source
turno
tec
met
Error: MS(Error)

DF

Type III SS

3
3
3

Mean Square

467.187500
17.187500
145.687500
6

22.875000

F Value

Pr > F

155.729167
40.85 0.0002
5.729167
1.50 0.3065
48.562500
12.74 0.0052
3.812500

The SAS System

09:32 Wednesday, October 20, 2014 5

Least Squares Means


met
1
2
3
4

tiempo
LSMEAN
LSMEAN
Number
90.2500000
94.7500000
96.2500000
98.5000000

1
2
3
4

Least Squares Means for Effect met


t for H0: LSMean(i)=LSMean(j) / Pr > |t|
Dependent Variable: tiempo
i/j

1
1
2
3
4

-3.25929
-4.34572
-5.97536
0.0173
0.0048
0.0010
3.259287
-1.08643
-2.71607
0.0173
0.3190
0.0348
4.345716
1.086429
-1.62964
0.0048
0.3190
0.1543
5.975359
2.716072
1.629643
0.0010
0.0348
0.1543

NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should
be used.

The SAS System

09:32 Wednesday, October 20, 2014 6

Dependent Variable: tiempo


Contrast

DF

1-2

Contrast SS
40.50000000

Mean Square
40.50000000

Standard
Estimate
Error

Parameter
1-2

4.50000000

t Value

1.38067013

3.26

proc mixed;
class turno tec met;
model tiempo=met/ddfm=satterth;
lsmeans met;
contrast met -1 1 0 0;
estimate met -1 1 0 0;
run;

The SAS System

09:32 Wednesday, October 20, 2014 7


The GLM Procedure
Class Level Information
Class
turno
tec
met

Levels
4
4
4

Values
1234

1234
1234

Number of observations

16

F Value
10.62

Pr > F
0.0173

Pr > |t|
0.0173

The SAS System

09:32 Wednesday, October 20, 2014 8

The GLM Procedure


Dependent Variable: tiempo
Source

DF

Sum of
Squares

Model

630.0625000

Error

Corrected Total

15

R-Square

turno
tec
met

3
3
3

18.36

Pr > F
0.0011

3.8125000

Root MSE

2.056682
DF

70.0069444

F Value

652.9375000

Coeff Var

0.964966
Source

22.8750000

Mean Square

1.952562

Type III SS
467.1875000
17.1875000
145.6875000

tiempo Mean
94.93750

Mean Square

F Value

Pr > F

155.7291667
40.85 0.0002
5.7291667
1.50 0.3065
48.5625000
12.74 0.0052

The SAS System

09:32 Wednesday, October 20, 2014 9

The GLM Procedure


Source
turno

Type III Expected Mean Square


Var(Error) + 4 Var(turno)

tec

Var(Error) + 4 Var(tec)

met

Var(Error) + Q(met)

The SAS System

09:32 Wednesday, October 20, 2014 10

The GLM Procedure


Tests of Hypotheses for Mixed Model Analysis of Variance
Dependent Variable: tiempo
Source
turno
tec
met
Error: MS(Error)

DF

Type III SS

3
3
3

Mean Square

467.187500
17.187500
145.687500
6

22.875000

F Value

Pr > F

155.729167
40.85 0.0002
5.729167
1.50 0.3065
48.562500
12.74 0.0052
3.812500

The SAS System

09:32 Wednesday, October 20, 2014 11

Least Squares Means


met
1
2
3
4

tiempo
LSMEAN
LSMEAN
Number
90.2500000
94.7500000
96.2500000
98.5000000

1
2
3
4

Least Squares Means for Effect met


t for H0: LSMean(i)=LSMean(j) / Pr > |t|
Dependent Variable: tiempo
i/j

1
1
2
3
4

-3.25929
-4.34572
-5.97536
0.0173
0.0048
0.0010
3.259287
-1.08643
-2.71607
0.0173
0.3190
0.0348
4.345716
1.086429
-1.62964
0.0048
0.3190
0.1543
5.975359
2.716072
1.629643
0.0010
0.0348
0.1543

NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should
be used.

The SAS System

09:32 Wednesday, October 20, 2014 12

Dependent Variable: tiempo


Contrast
1-2

Parameter
1-2

DF
1

Contrast SS
40.50000000

Mean Square
40.50000000

Standard
Estimate
Error
4.50000000

t Value

1.38067013

3.26

F Value
10.62

Pr > F
0.0173

Pr > |t|
0.0173

The SAS System

09:32 Wednesday, October 20, 2014 13

The GLM Procedure


Class Level Information
Class
turno
tec
met

Levels
4
4
4

Values
1234

1234
1234

Number of observations

16

The SAS System

09:32 Wednesday, October 20, 2014 14

The GLM Procedure


Dependent Variable: tiempo
Source

DF

Sum of
Squares

Model

630.0625000

Error

Corrected Total

15

R-Square

turno
tec
met

3
3
3

18.36

Pr > F
0.0011

3.8125000

Root MSE

2.056682
DF

70.0069444

F Value

652.9375000

Coeff Var

0.964966
Source

22.8750000

Mean Square

1.952562

Type III SS
467.1875000
17.1875000
145.6875000

tiempo Mean
94.93750

Mean Square

F Value

Pr > F

155.7291667
40.85 0.0002
5.7291667
1.50 0.3065
48.5625000
12.74 0.0052

The SAS System

09:32 Wednesday, October 20, 2014 15

The GLM Procedure


Source
turno

Type III Expected Mean Square


Var(Error) + 4 Var(turno)

tec

Var(Error) + 4 Var(tec)

met

Var(Error) + Q(met)

The SAS System

09:32 Wednesday, October 20, 2014 16

The GLM Procedure


Tests of Hypotheses for Mixed Model Analysis of Variance
Dependent Variable: tiempo
Source
turno
tec
met
Error: MS(Error)

DF

Type III SS

3
3
3

Mean Square

467.187500
17.187500
145.687500
6

22.875000

F Value

Pr > F

155.729167
40.85 0.0002
5.729167
1.50 0.3065
48.562500
12.74 0.0052
3.812500

The SAS System

09:32 Wednesday, October 20, 2014 17

Least Squares Means


met
1
2
3
4

tiempo
LSMEAN
LSMEAN
Number
90.2500000
94.7500000
96.2500000
98.5000000

1
2
3
4

Least Squares Means for Effect met


t for H0: LSMean(i)=LSMean(j) / Pr > |t|
Dependent Variable: tiempo
i/j

1
1
2
3
4

-3.25929
-4.34572
-5.97536
0.0173
0.0048
0.0010
3.259287
-1.08643
-2.71607
0.0173
0.3190
0.0348
4.345716
1.086429
-1.62964
0.0048
0.3190
0.1543
5.975359
2.716072
1.629643
0.0010
0.0348
0.1543

NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should
be used.

The SAS System

09:32 Wednesday, October 20, 2014 18

Dependent Variable: tiempo


Contrast
1-2

Parameter
1-2

DF
1

Contrast SS
40.50000000

Mean Square
40.50000000

Standard
Estimate
Error
4.50000000

t Value

1.38067013

3.26

F Value
10.62

Pr > F
0.0173

Pr > |t|
0.0173

The SAS System

09:32 Wednesday, October 20, 2014 19

The GLM Procedure


Class Level Information
Class
turno
tec
met

Levels
4
4
4

Values
1234

1234
1234

Number of observations

16

The SAS System

09:32 Wednesday, October 20, 2014 20

The GLM Procedure


Dependent Variable: tiempo
Source

DF

Sum of
Squares

Model

630.0625000

Error

Corrected Total

15

R-Square

turno
tec
met

3
3
3

18.36

Pr > F
0.0011

3.8125000

Root MSE

2.056682
DF

70.0069444

F Value

652.9375000

Coeff Var

0.964966
Source

22.8750000

Mean Square

1.952562

Type III SS
467.1875000
17.1875000
145.6875000

tiempo Mean
94.93750

Mean Square

F Value

Pr > F

155.7291667
40.85 0.0002
5.7291667
1.50 0.3065
48.5625000
12.74 0.0052

The SAS System

09:32 Wednesday, October 20, 2014 21

The GLM Procedure


Source
turno

Type III Expected Mean Square


Var(Error) + 4 Var(turno)

tec

Var(Error) + 4 Var(tec)

met

Var(Error) + Q(met)

The SAS System

09:32 Wednesday, October 20, 2014 22

The GLM Procedure


Tests of Hypotheses for Mixed Model Analysis of Variance
Dependent Variable: tiempo
Source
turno
tec
met
Error: MS(Error)

DF

Type III SS

3
3
3

Mean Square

467.187500
17.187500
145.687500
6

22.875000

F Value

Pr > F

155.729167
40.85 0.0002
5.729167
1.50 0.3065
48.562500
12.74 0.0052
3.812500

The SAS System

09:32 Wednesday, October 20, 2014 23

Least Squares Means


met
1
2
3
4

tiempo
LSMEAN
LSMEAN
Number
90.2500000
94.7500000
96.2500000
98.5000000

1
2
3
4

Least Squares Means for Effect met


t for H0: LSMean(i)=LSMean(j) / Pr > |t|
Dependent Variable: tiempo
i/j

1
1
2
3
4

-3.25929
-4.34572
-5.97536
0.0173
0.0048
0.0010
3.259287
-1.08643
-2.71607
0.0173
0.3190
0.0348
4.345716
1.086429
-1.62964
0.0048
0.3190
0.1543
5.975359
2.716072
1.629643
0.0010
0.0348
0.1543

NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should
be used.

The SAS System

09:32 Wednesday, October 20, 2014 24

Dependent Variable: tiempo


Contrast
1-2

Parameter
1-2

DF
1

Contrast SS
40.50000000

Mean Square
40.50000000

Standard
Estimate
Error
4.50000000

t Value

1.38067013

3.26

F Value
10.62

Pr > F
0.0173

Pr > |t|
0.0173

The SAS System

09:32 Wednesday, October 20, 2014 25

The Mixed Procedure


Model Information
Data Set
WORK.CL
Dependent Variable
tiempo
Covariance Structure
Variance Components
Estimation Method
REML
Residual Variance Method
Profile
Fixed Effects SE Method
Model-Based
Degrees of Freedom Method Satterthwaite
Class Level Information
Class

Levels

turno
tec
met

Values

4 1234
4 1234
4 1234
Dimensions

Covariance Parameters
3
Columns in X
5
Columns in Z
8
Subjects
1
Max Obs Per Subject
16
Observations Used
16
Observations Not Used
0
Total Observations
16
Iteration History
Iteration

Evaluations

0
1

1
1

-2 Res Log Like

84.52887019
68.01047852

0.00000000

Convergence criteria met.


Covariance Parameter
Estimates
Cov Parm
turno
tec
Residual

Estimate
37.9792
0.4792
3.8125

Fit Statistics
-2 Res Log Likelihood
AIC (smaller is better)
AICC (smaller is better)
BIC (smaller is better)

68.0
74.0
77.0
72.2

Type 3 Tests of Fixed Effects


Effect

Num
DF

Den
DF

F Value

Criterion

Pr > F

met

12.74

0.0052

Estimates
Label

Standard
Estimate
Error

1-2

4.5000

DF

t Value

Pr > |t|

3.26

0.0173

1.3807
Contrasts

Label

Num
DF

1-2

Den
DF

F Value
10.62

Pr > F

0.0173

Least Squares Means


Effect

met

met
met
met
met

1
2
3
4

Standard
Estimate
Error
90.2500
94.7500
96.2500
98.5000

3.2508
3.2508
3.2508
3.2508

DF
3.53
3.53
3.53
3.53

t Value
27.76
29.15
29.61
30.30

Pr > |t|
<.0001
<.0001
<.0001
<.0001

tarea:
The SAS System

09:32 Wednesday, October 20, 2014 26


The GLM Procedure
Class Level Information
Class

Levels

Values

grad

1234

pen

1234

trat

1234

Number of observations

16

The SAS System

09:32 Wednesday, October 20, 2014 27

The GLM Procedure


Dependent Variable: rend
Source

DF

Sum of
Squares

Model

251.2856250

Error

Corrected Total

35.4887500
15

R-Square

grad
pen
trat

3
3
3

27.9206250

4.72

Pr > F
0.0362

5.9147917

Root MSE

9.306996
DF

F Value

286.7743750

Coeff Var

0.876249
Source

Mean Square

2.432034

Type III SS
2.4068750
193.0668750
55.8118750

rend Mean
26.13125

Mean Square
0.8022917
64.3556250
18.6039583

F Value

Pr > F

0.14 0.9352
10.88 0.0077
3.15 0.1080

The SAS System

09:32 Wednesday, October 20, 2014 28

The GLM Procedure


Source

Type III Expected Mean Square

grad

Var(Error) + 4 Var(grad)

pen

Var(Error) + 4 Var(pen)

trat

Var(Error) + Q(trat)

The SAS System

09:32 Wednesday, October 20, 2014 29

The GLM Procedure


Tests of Hypotheses for Mixed Model Analysis of Variance
Dependent Variable: rend
Source
grad
pen
trat
Error: MS(Error)

DF
3
3
3
6

Type III SS

Mean Square

F Value

2.406875
193.066875
55.811875

0.802292
64.355625
18.603958

0.14 0.9352
10.88 0.0077
3.15 0.1080

35.488750

5.914792

Pr > F

The SAS System

09:32 Wednesday, October 20, 2014 30

Least Squares Means


trat

LSMEAN
rend LSMEAN
Number

1
2
3
4

24.7000000
29.3250000
25.5250000
24.9750000

1
2
3
4

Least Squares Means for Effect trat


t for H0: LSMean(i)=LSMean(j) / Pr > |t|
Dependent Variable: rend
i/j

1
1
2
3
4

-2.68941
-0.47973
-0.15991
0.0361
0.6484
0.8782
2.68941
2.209677
2.529499
0.0361
0.0692
0.0447
0.479733
-2.20968
0.319822
0.6484
0.0692
0.7599
0.159911
-2.5295
-0.31982
0.8782
0.0447
0.7599

NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should
be used.

options ls=120 ps=1000;


data cl;
input grad pen trat$ rend;
cards;
1
1
3
26.7
1
2
1
19.7
1
3
2
28
1
4
4
29.8
2
1
1
23.1
2
2
2
25.2
2
3
4
24.9
2
4
3
29
3
1
2
30
3
2
4
20.1
3
3
3
29.0
3
4
1
27.3
4
1
4
25.1
4
2
3
17.4
4
3
1
28.7
4
4
2
34.1
;
proc glm;
class grad pen trat;
model rend=grad pen trat/ss3;
random grad pen/test;
lsmeans trat/tdiff;
contrast '2-3' trat 0 1 -1 0;
estimate '2-3' trat 0 1 -1 0;
run;
proc mixed;
class grad pen trat;
model rend=trat/ddfm=satterth;
lsmeans trat;
contrast '2-3' trat 0 1 -1 0;
estimate '2-3' trat 0 1 -1 0;
run;
The SAS System

09:32 Wednesday, October 20, 2014 32


The GLM Procedure
Class Level Information

Class

Levels

Values

grad

1234

pen

1234

trat

1234

Number of observations

16

The SAS System

09:32 Wednesday, October 20, 2014 33

The GLM Procedure


Dependent Variable: rend
Source

DF

Sum of
Squares

Model

251.2856250

Error

Corrected Total

35.4887500
15

R-Square

grad
pen
trat

3
3
3

27.9206250

4.72

Pr > F
0.0362

5.9147917

Root MSE

9.306996
DF

F Value

286.7743750

Coeff Var

0.876249
Source

Mean Square

2.432034

Type III SS
2.4068750
193.0668750
55.8118750

rend Mean
26.13125

Mean Square
0.8022917
64.3556250
18.6039583

F Value

Pr > F

0.14 0.9352
10.88 0.0077
3.15 0.1080

The SAS System

09:32 Wednesday, October 20, 2014 34

The GLM Procedure


Source

Type III Expected Mean Square

grad

Var(Error) + 4 Var(grad)

pen

Var(Error) + 4 Var(pen)

trat

Var(Error) + Q(trat)

The SAS System

09:32 Wednesday, October 20, 2014 35

The GLM Procedure


Tests of Hypotheses for Mixed Model Analysis of Variance
Dependent Variable: rend
Source
grad
pen
trat
Error: MS(Error)

DF
3
3
3
6

Type III SS

Mean Square

F Value

2.406875
193.066875
55.811875

0.802292
64.355625
18.603958

0.14 0.9352
10.88 0.0077
3.15 0.1080

35.488750

5.914792

Pr > F

The SAS System

09:32 Wednesday, October 20, 2014 36

Least Squares Means


trat

LSMEAN
rend LSMEAN
Number

1
2
3
4

24.7000000
29.3250000
25.5250000
24.9750000

1
2
3
4

Least Squares Means for Effect trat


t for H0: LSMean(i)=LSMean(j) / Pr > |t|
Dependent Variable: rend
i/j

1
1
2
3
4

-2.68941
-0.47973
-0.15991
0.0361
0.6484
0.8782
2.68941
2.209677
2.529499
0.0361
0.0692
0.0447
0.479733
-2.20968
0.319822
0.6484
0.0692
0.7599
0.159911
-2.5295
-0.31982
0.8782
0.0447
0.7599

NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should
be used.

The SAS System

09:32 Wednesday, October 20, 2014 37

Dependent Variable: rend


Contrast
2-3

Parameter
2-3

DF
1

Contrast SS
28.88000000

Mean Square
28.88000000

Standard
Estimate
Error
3.80000000

t Value

1.71970807

2.21

F Value
4.88

Pr > F
0.0692

Pr > |t|
0.0692

The SAS System

09:32 Wednesday, October 20, 2014 38

The Mixed Procedure


Model Information
Data Set
WORK.CL
Dependent Variable
rend
Covariance Structure
Diagonal
Estimation Method
REML
Residual Variance Method
Profile
Fixed Effects SE Method
Model-Based
Degrees of Freedom Method Residual
Class Level Information
Class

Levels

grad
pen
trat

Values

4 1234
4 1234
4 1234
Dimensions

Covariance Parameters
1
Columns in X
5
Columns in Z
0
Subjects
1
Max Obs Per Subject
16
Observations Used
16
Observations Not Used
0
Total Observations
16
Covariance Parameter
Estimates
Cov Parm

Estimate

Residual

19.2469

Fit Statistics
-2 Res Log Likelihood
AIC (smaller is better)
AICC (smaller is better)
BIC (smaller is better)

75.1
77.1
77.5
77.6

Type 3 Tests of Fixed Effects


Effect

Num
DF

trat

Den
DF

F Value

12

0.97

0.4402

Pr > F

Estimates
Label

Standard
Estimate
Error

DF

t Value

2-3

3.8000

12

1.22

3.1022
Contrasts

Pr > |t|
0.2441

Label
2-3

Num
DF
1

Den
DF

F Value

12

1.50

Pr > F

0.2441

Least Squares Means


Effect
trat
trat
trat
trat

trat
1
2
3
4

Standard
Estimate
Error
24.7000
29.3250
25.5250
24.9750

2.1936
2.1936
2.1936
2.1936

The SAS System

DF

t Value

12
12
12
12

11.26
13.37
11.64
11.39

Model Information
Data Set
WORK.CL
Dependent Variable
rend
Covariance Structure
Diagonal
Estimation Method
REML
Residual Variance Method
Profile
Fixed Effects SE Method
Model-Based
Degrees of Freedom Method Residual
Class Level Information

pen
trat

Levels

<.0001
<.0001
<.0001
<.0001

09:32 Wednesday, October 20, 2014 39

The Mixed Procedure

Class

Pr > |t|

Values

4 1234
4 1234
Dimensions
Covariance Parameters
1
Columns in X
5
Columns in Z
0
Subjects
1
Max Obs Per Subject
16
Observations Used
16
Observations Not Used
0
Total Observations
16
Covariance Parameter
Estimates
Cov Parm

Estimate

Residual

19.2469

Fit Statistics
-2 Res Log Likelihood
AIC (smaller is better)
AICC (smaller is better)
BIC (smaller is better)

75.1
77.1
77.5
77.6

Type 3 Tests of Fixed Effects


Num
DF

Effect
trat

Den
DF

F Value

12

0.97

0.4402

Pr > F

Estimates
Label

Standard
Estimate
Error

DF

t Value

2-3

3.8000

12

1.22

3.1022

Pr > |t|
0.2441

Contrasts
Label
2-3

Num
DF
1

Den
DF

F Value

12

1.50

Pr > F

0.2441

Least Squares Means


Effect
trat
trat
trat
trat

trat
1
2
3
4

Standard
Estimate
Error
24.7000
29.3250
25.5250
24.9750

2.1936
2.1936
2.1936
2.1936

DF

t Value

12
12
12
12

11.26
13.37
11.64
11.39

Pr > |t|
<.0001
<.0001
<.0001
<.0001

También podría gustarte