EXST7015 : Statistical Techniques II Geaghan Random coefficients regression Weight Lifting example Page 1 1 /*---------------------------------------------------------------- 2 SAS System for Mixed Models (1996) 3 by Ramon C. Littell, Ph.D., George A. Milliken, Ph.D., 4 Walter W. Stroup, Ph.D., and Russell D. Wolfinger, Ph.D.(page 87 ff) 5 6 SAS Publications order # 55235 7 ISBN 1-55544-779-1 8 Copyright 1996 by SAS Institute Inc., Cary, NC, USA 9 10 This file contains the SAS code needed to produce the output 11 in this book with Release 6.11 of the SAS System. 12 ----------------------------------------------------------------*/ 13 14 /* Updated: 12JUN02 */ 15 /* Modifications by Geaghan 14Nov02 */ 16 17 ODS HTML style=minimal rs=none 18 body='C:\Geaghan\EXST\EXST7034New\Fall2002\SAS\Datasets Mixed 18 ! Model\MixModProgCh3.html' ; NOTE: Writing HTML Body file: C:\Geaghan\EXST\EXST7034New\Fall2002\SAS\MixModProgCh3.html 19 20 options ps=256 ls=100 nocenter nodate nonumber; 21 title1 'Random Coeff Reg - SAS System for Mixed Models, 1996, Littell, et al.'; 22 data weights; 23 input subj program$ s1 s2 s3 s4 s5 s6 s7; 24 datalines; NOTE: The data set WORK.WEIGHTS has 57 observations and 9 variables. NOTE: DATA statement used: real time 0.04 seconds 82 ; 83 84 data weight2; 85 set weights; 86 time=1; strength=s1; output; 87 time=2; strength=s2; output; 88 time=3; strength=s3; output; 89 time=4; strength=s4; output; 90 time=5; strength=s5; output; 91 time=6; strength=s6; output; 92 time=7; strength=s7; output; 93 * label subj = subject (experimental unit); 94 * label program = weight lifting program; 95 keep subj program time strength; 96 run; NOTE: There were 57 observations read from the data set WORK.WEIGHTS. NOTE: The data set WORK.WEIGHT2 has 399 observations and 4 variables. NOTE: DATA statement used: real time 0.05 seconds 97 98 99 data weight2; set weight2; t=time; tt=t*t; run; NOTE: There were 399 observations read from the data set WORK.WEIGHT2. NOTE: The data set WORK.WEIGHT2 has 399 observations and 6 variables. NOTE: DATA statement used: real time 0.00 seconds 100 101 title2 'ANCOVA without Random coefficients'; 102 proc mixed data=weight2; class program subj; 103 model strength = program time time*program time*time time*time*program / htype=1 3; 104 run; NOTE: The PROCEDURE MIXED printed page 1. NOTE: PROCEDURE MIXED used: real time 0.11 secondsEXST7015 : Statistical Techniques II Geaghan Random coefficients regression Weight Lifting example Page 2 Random Coeff Reg - SAS System for Mixed Models, 1996, Littell, et al. ANCOVA without Random coefficients The Mixed Procedure Model Information Data Set WORK.WEIGHT2 Dependent Variable strength 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 program 3 CONT RI WI subj 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Dimensions Covariance Parameters 1 Columns in X 12 Columns in Z 0 Subjects 1 Max Obs Per Subject 399 Observations Used 399 Observations Not Used 0 Total Observations 399 Covariance Parameter Estimates Cov Parm Estimate Residual 10.4770 Fit Statistics -2 Res Log Likelihood 2078.5 AIC (smaller is better) 2080.5 AICC (smaller is better) 2080.5 BIC (smaller is better) 2084.5 Type 1 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F program 2 390 20.02 <.0001 time 1 390 3.89 0.0493 time*program 2 390 1.93 0.1469 time*time 1 390 0.94 0.3332 time*time*program 2 390 0.07 0.9343 Type 3 Tests of Fixed Effects Effect Num DF Den DF F Value Pr > F program 2 390 0.39 0.6749 time 1 390 1.95 0.1629 time*program 2 390 0.19 0.8311 time*time 1 390 1.01 0.3156 time*time*program 2 390 0.07 0.9343 106 title2 'Random coefficient Regression'; 107 proc mixed data=weight2 covtest; class program subj; 108 model strength = program time time*program time*time time*time*program / htype=1 3; 109 random subj(program) time*subj(program) time*time*subj(program); 110 run; NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed page 2. NOTE: PROCEDURE MIXED used: real time 47.88 seconds 7980818283841234567EXST7015 : Statistical Techniques II Geaghan Random coefficients regression Weight Lifting example Page 3 Random Coeff Reg - SAS System for Mixed Models, 1996, Littell, et al. Random coefficient Regression The Mixed Procedure Model Information Data Set WORK.WEIGHT2 Dependent Variable strength Covariance Structure Variance Components Subject Effect subj Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model-Based Degrees of Freedom Method Containment Class Level Information Class Levels Values program 3 CONT RI WI subj 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Dimensions Covariance Parameters 4 Columns in X 12 Columns in Z 171 Subjects 1 Max Obs Per Subject 399 Observations Used 399 Observations Not Used 0 Total Observations 399 Iteration History Iteration Evaluations -2 Res Log Like Criterion 0 1 2078.51675124 1 2 1319.94926364 0.00485216 2