Chapter 5 : Mouse feeding example
Analysis of Variance with PROC MIXED

The Mixed Procedure

Model Information
Data Set WORK.MOUSEDIET
Dependent Variable LIFETIME
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 Values
DIET 6 N/N85 N/R40 N/R50 NP R/R50 lopro

Dimensions
Covariance Parameters 1
Columns in X 7
Columns in Z 0
Subjects 1
Max Obs Per Subject 349

Number of Observations
Number of Observations Read 349
Number of Observations Used 349
Number of Observations Not Used 0

Covariance Parameter Estimates
Cov Parm Estimate Standard Error Z Value Pr Z Alpha Lower Upper
Residual 44.5989 3.4056 13.10 <.0001 0.05 38.6102 52.1055

Fit Statistics
-2 Res Log Likelihood 2300.3
AIC (smaller is better) 2302.3
AICC (smaller is better) 2302.4
BIC (smaller is better) 2306.2

Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
DIET 5 343 57.10 <.0001


Chapter 5 : Mouse feeding example
Testing homogeniety of Variance with PROC MIXED

The Mixed Procedure

Model Information
Data Set WORK.MOUSEDIET
Dependent Variable LIFETIME
Covariance Structure Variance Components
Group Effect DIET
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Prasad-Rao-Jeske-Kackar-Harville
Degrees of Freedom Method Kenward-Roger

Class Level Information
Class Levels Values
DIET 6 N/N85 N/R40 N/R50 NP R/R50 lopro

Dimensions
Covariance Parameters 6
Columns in X 7
Columns in Z 0
Subjects 349
Max Obs Per Subject 1

Number of Observations
Number of Observations Read 349
Number of Observations Used 349
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 2300.34852165  
1 1 2289.27652629 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Group Estimate Standard Error Z Value Pr Z Alpha Lower Upper
Residual DIET N/N85 26.2687 4.9643 5.29 <.0001 0.05 18.7234 39.5319
Residual DIET N/R40 44.9356 8.2733 5.43 <.0001 0.05 32.2855 66.8452
Residual DIET N/R50 60.3448 10.2001 5.92 <.0001 0.05 44.4538 86.6356
Residual DIET NP 37.6223 7.6796 4.90 <.0001 0.05 26.1635 58.7189
Residual DIET R/R50 44.6645 8.5172 5.24 <.0001 0.05 31.7464 67.4911
Residual DIET lopro 48.8838 9.3218 5.24 <.0001 0.05 34.7453 73.8667

Fit Statistics
-2 Res Log Likelihood 2289.3
AIC (smaller is better) 2301.3
AICC (smaller is better) 2301.5
BIC (smaller is better) 2324.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
5 11.07 0.0500

Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
DIET 5 155 64.71 <.0001


Chapter 5 : Mouse feeding example
Testing homogeniety of Variance with PROC MIXED

The Mixed Procedure

Model Information
Data Set WORK.MOUSEDIET
Dependent Variable LIFETIME
Covariance Structure Variance Components
Group Effect DIET
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Prasad-Rao-Jeske-Kackar-Harville
Degrees of Freedom Method Kenward-Roger

Class Level Information
Class Levels Values
DIET 6 N/N85 N/R40 N/R50 NP R/R50 lopro

Dimensions
Covariance Parameters 6
Columns in X 7
Columns in Z 0
Subjects 349
Max Obs Per Subject 1

Number of Observations
Number of Observations Read 349
Number of Observations Used 349
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 2300.34852165  
1 1 2289.27652629 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Group Estimate Standard Error Z Value Pr Z Alpha Lower Upper
Residual DIET N/N85 26.2687 4.9643 5.29 <.0001 0.05 18.7234 39.5319
Residual DIET N/R40 44.9356 8.2733 5.43 <.0001 0.05 32.2855 66.8452
Residual DIET N/R50 60.3448 10.2001 5.92 <.0001 0.05 44.4538 86.6356
Residual DIET NP 37.6223 7.6796 4.90 <.0001 0.05 26.1635 58.7189
Residual DIET R/R50 44.6645 8.5172 5.24 <.0001 0.05 31.7464 67.4911
Residual DIET lopro 48.8838 9.3218 5.24 <.0001 0.05 34.7453 73.8667

Fit Statistics
-2 Res Log Likelihood 2289.3
AIC (smaller is better) 2301.3
AICC (smaller is better) 2301.5
BIC (smaller is better) 2324.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
5 11.07 0.0500

Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
DIET 5 155 64.71 <.0001

Contrasts
Label Num DF Den DF F Value Pr > F
A: N/N85 N/R50 1 122 70.40 <.0001
B: N/R50 R/R50 1 124 0.21 0.6474
C: N/R40 N/R50 1 129 4.97 0.0275
D: N/R50 lopro 1 123 3.96 0.0489
E: N/N85 NP 1 93.9 22.77 <.0001


Chapter 5 : Mouse feeding example
Analysis wtih PROC GLM

The GLM Procedure

Class Level Information
Class Levels Values
DIET 6 N/N85 N/R40 N/R50 NP R/R50 lopro

Number of Observations Read 349
Number of Observations Used 349


Chapter 5 : Mouse feeding example
Analysis wtih PROC GLM

The GLM Procedure
 
Dependent Variable: LIFETIME

Source DF Sum of Squares Mean Square F Value Pr > F
Model 5 12733.94181 2546.78836 57.10 <.0001
Error 343 15297.41532 44.59888    
Corrected Total 348 28031.35713      

R-Square Coeff Var Root MSE LIFETIME Mean
0.454275 17.21323 6.678239 38.79713

Source DF Type I SS Mean Square F Value Pr > F
DIET 5 12733.94181 2546.78836 57.10 <.0001

Source DF Type III SS Mean Square F Value Pr > F
DIET 5 12733.94181 2546.78836 57.10 <.0001

Contrast DF Contrast SS Mean Square F Value Pr > F
A: N/N85 N/R50 1 2917.456121 2917.456121 65.42 <.0001
B: N/R50 R/R50 1 10.843803 10.843803 0.24 0.6223
C: N/R40 N/R50 1 258.510054 258.510054 5.80 0.0166
D: N/R50 lopro 1 213.507110 213.507110 4.79 0.0293
E: N/N85 NP 1 737.128081 737.128081 16.53 <.0001


Chapter 5 : Mouse feeding example
Analysis wtih PROC GLM

The Mixed Procedure

Model Information
Data Set WORK.MOUSEDIET
Dependent Variable LIFETIME
Covariance Structure Variance Components
Group Effect DIET
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Prasad-Rao-Jeske-Kackar-Harville
Degrees of Freedom Method Kenward-Roger

Class Level Information
Class Levels Values
DIET 6 N/N85 N/R40 N/R50 NP R/R50 lopro

Dimensions
Covariance Parameters 6
Columns in X 7
Columns in Z 0
Subjects 349
Max Obs Per Subject 1

Number of Observations
Number of Observations Read 349
Number of Observations Used 349
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 2300.34852165  
1 1 2289.27652629 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Group Estimate Standard Error Z Value Pr Z Alpha Lower Upper
Residual DIET N/N85 26.2687 4.9643 5.29 <.0001 0.05 18.7234 39.5319
Residual DIET N/R40 44.9356 8.2733 5.43 <.0001 0.05 32.2855 66.8452
Residual DIET N/R50 60.3448 10.2001 5.92 <.0001 0.05 44.4538 86.6356
Residual DIET NP 37.6223 7.6796 4.90 <.0001 0.05 26.1635 58.7189
Residual DIET R/R50 44.6645 8.5172 5.24 <.0001 0.05 31.7464 67.4911
Residual DIET lopro 48.8838 9.3218 5.24 <.0001 0.05 34.7453 73.8667

Fit Statistics
-2 Res Log Likelihood 2289.3
AIC (smaller is better) 2301.3
AICC (smaller is better) 2301.5
BIC (smaller is better) 2324.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
5 11.07 0.0500

Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
DIET 5 155 64.71 <.0001

Contrasts
Label Num DF Den DF F Value Pr > F
1 1 147 141.01 <.0001
2 1 147 141.01 <.0001
3 1 74.4 2.54 0.1154
4 1 74.4 2.54 0.1154
5 1 221 303.02 <.0001
6 1 221 303.02 <.0001
7 . . . .
8 1 236 312.50 <.0001


Chapter 5 : Mouse feeding example
Analysis of Variance with PROC MIXED

The Mixed Procedure

Model Information
Data Set WORK.MOUSEDIET
Dependent Variable LIFETIME
Covariance Structure Variance Components
Group Effect DIET
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Prasad-Rao-Jeske-Kackar-Harville
Degrees of Freedom Method Kenward-Roger

Class Level Information
Class Levels Values
DIET 6 N/N85 N/R40 N/R50 NP R/R50 lopro

Dimensions
Covariance Parameters 6
Columns in X 7
Columns in Z 0
Subjects 349
Max Obs Per Subject 1

Number of Observations
Number of Observations Read 349
Number of Observations Used 349
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 2300.34852165  
1 1 2289.27652629 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Group Estimate Standard Error Z Value Pr Z Alpha Lower Upper
Residual DIET N/N85 26.2687 4.9643 5.29 <.0001 0.05 18.7234 39.5319
Residual DIET N/R40 44.9356 8.2733 5.43 <.0001 0.05 32.2855 66.8452
Residual DIET N/R50 60.3448 10.2001 5.92 <.0001 0.05 44.4538 86.6356
Residual DIET NP 37.6223 7.6796 4.90 <.0001 0.05 26.1635 58.7189
Residual DIET R/R50 44.6645 8.5172 5.24 <.0001 0.05 31.7464 67.4911
Residual DIET lopro 48.8838 9.3218 5.24 <.0001 0.05 34.7453 73.8667

Fit Statistics
-2 Res Log Likelihood 2289.3
AIC (smaller is better) 2301.3
AICC (smaller is better) 2301.5
BIC (smaller is better) 2324.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
5 11.07 0.0500

Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
DIET 5 155 64.71 <.0001

Least Squares Means
Effect DIET Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
DIET N/N85 32.6912 0.6789 56 48.16 <.0001 0.05 31.3313 34.0512
DIET N/R40 45.1167 0.8654 59 52.13 <.0001 0.05 43.3850 46.8483
DIET N/R50 42.2972 0.9219 70 45.88 <.0001 0.05 40.4585 44.1359
DIET NP 27.4020 0.8762 48 31.27 <.0001 0.05 25.6402 29.1638
DIET R/R50 42.8857 0.8931 55 48.02 <.0001 0.05 41.0960 44.6755
DIET lopro 39.6857 0.9343 55 42.48 <.0001 0.05 37.8133 41.5581

Differences of Least Squares Means
Effect DIET _DIET Estimate Standard Error DF t Value Pr > |t| Adjustment Adj P Alpha Lower Upper Adj Lower Adj Upper
DIET N/N85 N/R40 -12.4254 1.0999 110 -11.30 <.0001 Tukey-Kramer <.0001 0.05 -14.6052 -10.2457 -15.5995 -9.2514
DIET N/N85 N/R50 -9.6060 1.1449 122 -8.39 <.0001 Tukey-Kramer <.0001 0.05 -11.8724 -7.3395 -12.9099 -6.3020
DIET N/N85 NP 5.2892 1.1084 93.9 4.77 <.0001 Tukey-Kramer <.0001 0.05 3.0883 7.4901 2.0905 8.4879
DIET N/N85 R/R50 -10.1945 1.1218 103 -9.09 <.0001 Tukey-Kramer <.0001 0.05 -12.4193 -7.9697 -13.4318 -6.9572
DIET N/N85 lopro -6.9945 1.1549 101 -6.06 <.0001 Tukey-Kramer <.0001 0.05 -9.2855 -4.7034 -10.3273 -3.6617
DIET N/R40 N/R50 2.8195 1.2645 129 2.23 0.0275 Tukey-Kramer 0.2302 0.05 0.3177 5.3213 -0.8295 6.4684
DIET N/R40 NP 17.7146 1.2316 106 14.38 <.0001 Tukey-Kramer <.0001 0.05 15.2728 20.1564 14.1606 21.2686
DIET N/R40 R/R50 2.2310 1.2436 113 1.79 0.0755 Tukey-Kramer 0.4727 0.05 -0.2327 4.6946 -1.3578 5.8197
DIET N/R40 lopro 5.4310 1.2735 113 4.26 <.0001 Tukey-Kramer 0.0005 0.05 2.9078 7.9541 1.7559 9.1061
DIET N/R50 NP 14.8951 1.2719 116 11.71 <.0001 Tukey-Kramer <.0001 0.05 12.3759 17.4143 11.2247 18.5656
DIET N/R50 R/R50 -0.5885 1.2836 124 -0.46 0.6474 Tukey-Kramer 0.9974 0.05 -3.1290 1.9520 -4.2926 3.1155
DIET N/R50 lopro 2.6115 1.3126 123 1.99 0.0489 Tukey-Kramer 0.3530 0.05 0.01326 5.2097 -1.1763 6.3993
DIET NP R/R50 -15.4837 1.2512 103 -12.38 <.0001 Tukey-Kramer <.0001 0.05 -17.9651 -13.0022 -19.0942 -11.8731
DIET NP lopro -12.2837 1.2809 103 -9.59 <.0001 Tukey-Kramer <.0001 0.05 -14.8241 -9.7433 -15.9801 -8.5873
DIET R/R50 lopro 3.2000 1.2925 110 2.48 0.0148 Tukey-Kramer 0.1378 0.05 0.6385 5.7615 -0.5298 6.9298