Table 8.1, page 275.
data sm;
set smallmice;
cont_day = day;
cont_day2 = day**2;
run;
*Model #7: ARH(1);
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2 /notest;
repeated day / subject = id type = arh(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Heterogeneous
Autoregressive
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 990.9
AIC (smaller is better) 1006.9
AICC (smaller is better) 1008.5
BIC (smaller is better) 1012.0
DF Chi-Square Pr > ChiSq
7 172.22 <.0001
*Model 12: ANTEH;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2 /notest;
repeated day / subject = id type = ante(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Ante-dependence
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 983.1
AIC (smaller is better) 1009.1
AICC (smaller is better) 1013.6
BIC (smaller is better) 1017.4
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
12 180.00 <.0001
*Model 13: FAH(2);
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = fah(2);
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Factor Analytic
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 977.1
AIC (smaller is better) 1015.1
AICC (smaller is better) 1025.2
BIC (smaller is better) 1027.2
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
18 185.98 <.0001
*Model 14: UN;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = un;
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Unstructured
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 964.7
AIC (smaller is better) 1020.7
AICC (smaller is better) 1045.3
BIC (smaller is better) 1038.6
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
27 198.41 <.0001
*Model 2: AR(1);
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = ar(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Autoregressive
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1035.3
AIC (smaller is better) 1039.3
AICC (smaller is better) 1039.4
BIC (smaller is better) 1040.5
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 127.83 <.0001
*Model 5: ARMA;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = arma(1,1);
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Autoregressive
Moving Average
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1032.9
AIC (smaller is better) 1038.9
AICC (smaller is better) 1039.2
BIC (smaller is better) 1040.8
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 130.18 <.0001
*Model 10: FA(2);
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = fa1(2);
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Factor Analytic
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1004.0
AIC (smaller is better) 1032.0
AICC (smaller is better) 1037.3
BIC (smaller is better) 1041.0
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
13 159.05 <.0001
*Model 8: FA(1);
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = fa1(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Factor Analytic
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1020.4
AIC (smaller is better) 1036.4
AICC (smaller is better) 1038.1
BIC (smaller is better) 1041.5
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
7 142.69 <.0001
*Model 3: RS;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
random cont_day/ subject = id type = un;
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Unstructured
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Containment
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1041.0
AIC (smaller is better) 1045.0
AICC (smaller is better) 1045.1
BIC (smaller is better) 1046.3
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 122.10 <.0001
*Model 6: RIAS;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
random intercept cont_day / subject = id type = un;
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Unstructured
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Containment
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1038.6
AIC (smaller is better) 1046.6
AICC (smaller is better) 1047.0
BIC (smaller is better) 1049.2
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
3 124.48 <.0001
*Model 11: FAH(1);
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = fa(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Factor Analytic
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1016.8
AIC (smaller is better) 1044.8
AICC (smaller is better) 1050.0
BIC (smaller is better) 1053.7
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
13 146.33 <.0001
*Model 9: CSH;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = csh;
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Heterogeneous Compound
Symmetry
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1050.7
AIC (smaller is better) 1066.7
AICC (smaller is better) 1068.3
BIC (smaller is better) 1071.8
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
7 112.43 <.0001
*Model 4: RI;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = cs;
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Compound Symmetry
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1105.3
AIC (smaller is better) 1109.3
AICC (smaller is better) 1109.4
BIC (smaller is better) 1110.6
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 57.80 <.0001
*Model 1: IND;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ notest;
repeated day / subject = id type = vc;
run;
The Mixed Procedure
Model Information
Data Set WORK.SM
Dependent Variable weight
Covariance Structure Variance Components
Subject Effect id
Estimation Method REML
Residual Variance Method Parameter
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 1163.1
AIC (smaller is better) 1165.1
AICC (smaller is better) 1165.1
BIC (smaller is better) 1165.7
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
0 0.00 1.0000
Table 8.3, page 277.
*arh(1);
proc mixed data = sm method = reml noitprint noclprint noinfo;
class id day;
model weight = cont_day cont_day2/ solution notest;
repeated day / subject = id type = arh(1) rcorr;
run;
The Mixed Procedure
Estimated R Correlation Matrix for id 22
Row Col1 Col2 Col3 Col4 Col5 Col6 Col7
1 1.0000 0.8963 0.8034 0.7201 0.6454 0.5785 0.5185
2 0.8963 1.0000 0.8963 0.8034 0.7201 0.6454 0.5785
3 0.8034 0.8963 1.0000 0.8963 0.8034 0.7201 0.6454
4 0.7201 0.8034 0.8963 1.0000 0.8963 0.8034 0.7201
5 0.6454 0.7201 0.8034 0.8963 1.0000 0.8963 0.8034
6 0.5785 0.6454 0.7201 0.8034 0.8963 1.0000 0.8963
7 0.5185 0.5785 0.6454 0.7201 0.8034 0.8963 1.0000
[...output omitted...]
*ANTEH;
proc mixed data = sm method = reml noitprint noclprint noinfo;
class id day;
model weight = cont_day cont_day2/ solution notest ;
repeated day / subject = id type = ante(1) rcorr;
run;
The Mixed Procedure
Estimated R Correlation Matrix for id 22
Row Col1 Col2 Col3 Col4 Col5 Col6 Col7
1 1.0000 0.9228 0.6901 0.5995 0.5578 0.5217 0.4787
2 0.9228 1.0000 0.7478 0.6496 0.6044 0.5653 0.5187
3 0.6901 0.7478 1.0000 0.8687 0.8082 0.7560 0.6937
4 0.5995 0.6496 0.8687 1.0000 0.9304 0.8703 0.7985
5 0.5578 0.6044 0.8082 0.9304 1.0000 0.9353 0.8582
6 0.5217 0.5653 0.7560 0.8703 0.9353 1.0000 0.9176
7 0.4787 0.5187 0.6937 0.7985 0.8582 0.9176 1.0000
[...output omitted...]
*FAH(2);
proc mixed data = sm method = reml noclprint noinfo noitprint;
class id day;
model weight = cont_day cont_day2/ solution notest ;
repeated day / subject = id type = fah(2) rcorr;
run;
The Mixed Procedure
Estimated R Correlation Matrix for id 22
Row Col1 Col2 Col3 Col4 Col5 Col6 Col7
1 1.0000 0.9276 0.5844 0.3110 0.2056 0.2189 0.3083
2 0.9276 1.0000 0.7424 0.4978 0.3998 0.4090 0.4886
3 0.5844 0.7424 1.0000 0.8311 0.8024 0.7946 0.8064
4 0.3110 0.4978 0.8311 1.0000 0.9304 0.9138 0.8849
5 0.2056 0.3998 0.8024 0.9304 1.0000 0.9416 0.8984
6 0.2189 0.4090 0.7946 0.9138 0.9416 1.0000 0.8825
7 0.3083 0.4886 0.8064 0.8849 0.8984 0.8825 1.0000
*UN;
proc mixed data = sm method = reml noitprint noclprint noinfo;
class id day;
model weight = cont_day cont_day2/ solution notest ;
repeated day / subject = id type = un rcorr;
run;
The Mixed Procedure
Estimated R Correlation Matrix for id 22
Row Col1 Col2 Col3 Col4 Col5 Col6 Col7
1 1.0000 0.9188 0.4962 0.3033 0.1785 0.2329 0.3489
2 0.9188 1.0000 0.7317 0.5060 0.4205 0.4178 0.5308
3 0.4962 0.7317 1.0000 0.8737 0.8136 0.7500 0.8105
4 0.3033 0.5060 0.8737 1.0000 0.9330 0.9010 0.8665
5 0.1785 0.4205 0.8136 0.9330 1.0000 0.9372 0.8871
6 0.2329 0.4178 0.7500 0.9010 0.9372 1.0000 0.9191
7 0.3489 0.5308 0.8105 0.8665 0.8871 0.9191 1.0000
Table 8.4, page 278.
*ARH(1);
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ solution notest;
repeated day / subject = id type = arh(1);
run;
[...output omitted...]
Solution for Fixed Effects
Standard
Effect Estimate Error DF t Value Pr > |t|
Intercept 78.8108 6.9230 13 11.38 <.0001
cont_day 67.1741 3.0942 82 21.71 <.0001
cont_day2 -1.1793 0.1421 82 -8.30 <.0001
*ANTE;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ solution notest ;
repeated day / subject = id type = ante(1);
run;
[...output omitted...]
Solution for Fixed Effects
Standard
Effect Estimate Error DF t Value Pr > |t|
Intercept 81.3278 6.1346 13 13.26 <.0001
cont_day 66.3829 2.3550 82 28.19 <.0001
cont_day2 -1.1773 0.1152 82 -10.22 <.0001
*FAH(2);
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ solution notest ;
repeated day / subject = id type = fah(2);
run;
[...output omitted...]
Solution for Fixed Effects
Standard
Effect Estimate Error DF t Value Pr > |t|
Intercept 84.7254 5.6912 13 14.89 <.0001
cont_day 67.1351 2.2533 82 29.79 <.0001
cont_day2 -1.2583 0.08510 82 -14.79 <.0001
*UN;
proc mixed data = sm method = reml noclprint noitprint;
class id day;
model weight = cont_day cont_day2/ solution notest ;
repeated day / subject = id type = un;
run;
[...output omitted...]
Standard
Effect Estimate Error DF t Value Pr > |t|
Intercept 89.5909 4.5755 13 19.58 <.0001
cont_day 64.1899 2.3275 13 27.58 <.0001
cont_day2 -1.1105 0.09463 13 -11.74 <.0001
Table 8.5, page 279.
data pain1;
set pain;
tmt = "Z";
if trial = 4 and treatment = "attend" then tmt = "A";
if trial = 4 and treatment = "distract" then tmt = "D";
if trial = 4 and treatment = "no directions" then tmt = "N";
cont_trial = trial;
logpain = log(paintol);
run;
*RI;
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
random intercept / subject = id type = un;
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Unstructured
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Containment
Dimensions
Covariance Parameters 2
Columns in X 15
Columns in Z Per Subject 1
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 408.2
AIC (smaller is better) 412.2
AICC (smaller is better) 412.3
BIC (smaller is better) 416.5
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 142.92 <.0001
*ARMA(1,1);
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = arma(1,1);
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Autoregressive
Moving Average
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 3
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 406.1
AIC (smaller is better) 412.1
AICC (smaller is better) 412.2
BIC (smaller is better) 418.6
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 145.01 <.0001
*RIAS;
proc mixed data = pain1 method = reml noclprint noitprint;
class id tmt cs;
model logpain = cs|tmt trial/ notest;
random intercept trial / subject = id type = un;
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Unstructured
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Containment
Dimensions
Covariance Parameters 4
Columns in X 16
Columns in Z Per Subject 2
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 408.0
AIC (smaller is better) 416.0
AICC (smaller is better) 416.1
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 146.10 <.0001
*FA(1);
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = fa1(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Factor Analytic
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 5
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 403.9
AIC (smaller is better) 413.9
AICC (smaller is better) 414.2
BIC (smaller is better) 424.7
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
4 147.23 <.0001
*UN;
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = un;
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Unstructured
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 10
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 385.1
AIC (smaller is better) 405.1
AICC (smaller is better) 406.1
BIC (smaller is better) 426.7
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 166.04 <.0001
*FA(2);
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = fa1(2);
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Factor Analytic
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 8
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 396.7
AIC (smaller is better) 412.7
AICC (smaller is better) 413.3
BIC (smaller is better) 429.9
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
7 154.46 <.0001
*FAH(1);
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = fa(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Factor Analytic
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 8
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 396.8
AIC (smaller is better) 412.8
AICC (smaller is better) 413.4
BIC (smaller is better) 430.0
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
7 154.39 <.0001
*AR(1);
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = ar(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Autoregressive
Subject Effect id
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 2
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 428.9
AIC (smaller is better) 432.9
AICC (smaller is better) 432.9
BIC (smaller is better) 437.2
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 122.29 <.0001
*ANTEH(1);
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = ante(1);
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Ante-dependence
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 7
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 426.3
AIC (smaller is better) 440.3
AICC (smaller is better) 440.8
BIC (smaller is better) 455.4
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 124.82 <.0001
*IND;
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = vc;
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Variance Components
Subject Effect id
Estimation Method REML
Residual Variance Method Parameter
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 1
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
[...output omitted...]
Fit Statistics
-2 Res Log Likelihood 551.1
AIC (smaller is better) 553.1
AICC (smaller is better) 553.2
BIC (smaller is better) 555.3
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
0 0.00 1.0000
Table 8.6, page 280.
proc mixed data = pain1 method = reml noclprint noitprint;
class id trial tmt cs;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = un rcorr;
run;
The Mixed Procedure
Model Information
Data Set WORK.PAIN1
Dependent Variable logpain
Covariance Structure Unstructured
Subject Effect id
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Between-Within
Dimensions
Covariance Parameters 10
Columns in X 15
Columns in Z 0
Subjects 64
Max Obs Per Subject 4
Number of Observations
Number of Observations Read 256
Number of Observations Used 245
Number of Observations Not Used 11
Estimated R Correlation Matrix for id 1
Row Col1 Col2 Col3 Col4
1 1.0000 0.7015 0.8196 0.5489
2 0.7015 1.0000 0.7072 0.6727
3 0.8196 0.7072 1.0000 0.7326
4 0.5489 0.6727 0.7326 1.0000
Covariance Parameter Estimates
Cov Parm Subject Estimate
UN(1,1) id 0.4847
UN(2,1) id 0.3593
UN(2,2) id 0.5411
UN(3,1) id 0.4564
UN(3,2) id 0.4161
UN(3,3) id 0.6398
UN(4,1) id 0.2764
UN(4,2) id 0.3578
UN(4,3) id 0.4238
UN(4,4) id 0.5230
Fit Statistics
-2 Res Log Likelihood 385.1
AIC (smaller is better) 405.1
AICC (smaller is better) 406.1
BIC (smaller is better) 426.7
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 166.04 <.0001
Table 8.7, page 281.
data pain2;
set pain;
tmt_a = (treatment = "attend")*(trial = 4);
tmt_d = (treatment = "distract")*(trial = 4);
tmt_n = (treatment = "no directions")*(trial = 4);
attend = (cs = "attender");
distract = (cs = "distracter");
logpain = log(paintol);
run;
*TOP MODEL;
proc mixed data = pain2 method = reml noclprint noitprint;
class id trial ;
model logpain = attend tmt_a tmt_d tmt_n attend*tmt_a
attend*tmt_d attend*tmt_n/ notest solution;
repeated trial / subject = id type = un;
estimate 'DA' tmt_a 1 ;
estimate 'DD' tmt_d 1 ;
estimate 'DN' tmt_n 1 ;
estimate 'AA' attend*tmt_a 1 tmt_a 1 ;
estimate 'AD' attend*tmt_d 1 tmt_d 1 ;
estimate 'AN' attend*tmt_n 1 tmt_n 1 ;
run;
The Mixed Procedure
[...output omitted...]
Solution for Fixed Effects
Standard
Effect Estimate Error DF t Value Pr > |t|
Intercept 3.5109 0.1167 62 30.08 <.0001
attend -0.3560 0.1650 62 -2.16 0.0348
tmt_a -0.1352 0.1467 62 -0.92 0.3603
tmt_d 0.4900 0.1403 62 3.49 0.0009
tmt_n -0.3495 0.1509 62 -2.32 0.0238
attend*tmt_a 0.2317 0.2075 62 1.12 0.2684
attend*tmt_d -0.4245 0.2030 62 -2.09 0.0406
attend*tmt_n 0.3313 0.2118 62 1.56 0.1228
Estimates
Standard
Label Estimate Error DF t Value Pr > |t|
DA -0.1352 0.1467 62 -0.92 0.3603
DD 0.4900 0.1403 62 3.49 0.0009
DN -0.3495 0.1509 62 -2.32 0.0238
AA 0.09650 0.1467 62 0.66 0.5131
AD 0.06547 0.1467 62 0.45 0.6569
AN -0.01820 0.1486 62 -0.12 0.9029
*Bottom MODEL;
proc mixed data = pain2 method = reml noclprint noitprint;
class id trial ;
model logpain = attend tmt_a tmt_d tmt_n attend*tmt_a attend*tmt_d attend*tmt_n/ notest solution;
random intercept / subject = id type = cs;
estimate 'DA' tmt_a 1 ;
estimate 'DD' tmt_d 1 ;
estimate 'DN' tmt_n 1 ;
estimate 'AA' attend*tmt_a 1 tmt_a 1 ;
estimate 'AD' attend*tmt_d 1 tmt_d 1 ;
estimate 'AN' attend*tmt_n 1 tmt_n 1 ;
run;
The Mixed Procedure
[...output omitted...]
Solution for Fixed Effects
Standard
Effect Estimate Error DF t Value Pr > |t|
Intercept 3.4731 0.1173 62 29.62 <.0001
attend -0.3313 0.1659 175 -2.00 0.0474
tmt_a -0.2438 0.1467 175 -1.66 0.0982
tmt_d 0.3993 0.1399 175 2.85 0.0048
tmt_n -0.3340 0.1493 175 -2.24 0.0266
attend*tmt_a 0.3188 0.2074 175 1.54 0.1261
attend*tmt_d -0.3723 0.2027 175 -1.84 0.0680
attend*tmt_n 0.2590 0.2098 175 1.23 0.2187
Estimates
Standard
Label Estimate Error DF t Value Pr > |t|
DA -0.2438 0.1467 175 -1.66 0.0982
DD 0.3993 0.1399 175 2.85 0.0048
DN -0.3340 0.1493 175 -2.24 0.0266
AA 0.07499 0.1467 175 0.51 0.6098
AD 0.02700 0.1467 175 0.18 0.8542
AN -0.07500 0.1474 175 -0.51 0.6114
Table 8.8, page 283.
*RI-2;
proc mixed data = pain1 method = reml noclprint noitprint noinfo;
class cs treatment id trial tmt;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = cs group = cs;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Group Estimate
Variance id cs attender 0.09966
CS id cs attender 0.3004
Variance id cs distracter 0.2294
CS id cs distracter 0.4630
Fit Statistics
-2 Res Log Likelihood 391.7
AIC (smaller is better) 399.7
AICC (smaller is better) 399.9
BIC (smaller is better) 408.4
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
3 159.40 <.0001
*UN-2;
proc mixed data = pain1 method = reml noclprint noitprint noinfo;
class cs tmt id trial;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = un group = cs;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Group Estimate
UN(1,1) id cs attender 0.3800
UN(2,1) id cs attender 0.3599
UN(2,2) id cs attender 0.4801
UN(3,1) id cs attender 0.3360
UN(3,2) id cs attender 0.3390
UN(3,3) id cs attender 0.3980
UN(4,1) id cs attender 0.2244
UN(4,2) id cs attender 0.2959
UN(4,3) id cs attender 0.2655
UN(4,4) id cs attender 0.3582
UN(1,1) id cs distracter 0.5874
UN(2,1) id cs distracter 0.3621
UN(2,2) id cs distracter 0.6100
UN(3,1) id cs distracter 0.5769
UN(3,2) id cs distracter 0.4978
UN(3,3) id cs distracter 0.8828
UN(4,1) id cs distracter 0.3323
UN(4,2) id cs distracter 0.4283
UN(4,3) id cs distracter 0.5878
UN(4,4) id cs distracter 0.6941
Fit Statistics
-2 Res Log Likelihood 361.2
AIC (smaller is better) 401.2
AICC (smaller is better) 405.1
BIC (smaller is better) 444.4
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
19 189.96 <.0001
*RI;
proc mixed data = pain1 method = reml noclprint noitprint noinfo;
class cs treatment id trial tmt;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = cs;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Estimate
CS id 0.3814
Residual 0.1649
Fit Statistics
-2 Res Log Likelihood 408.2
AIC (smaller is better) 412.2
AICC (smaller is better) 412.3
BIC (smaller is better) 416.5
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 142.92 <.0001
*UN;
proc mixed data = pain1 method = reml noclprint noitprint noinfo;
class cs tmt id trial;
model logpain = cs|tmt/ notest;
repeated trial / subject = id type = un;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Estimate
UN(1,1) id 0.4847
UN(2,1) id 0.3593
UN(2,2) id 0.5411
UN(3,1) id 0.4564
UN(3,2) id 0.4161
UN(3,3) id 0.6398
UN(4,1) id 0.2764
UN(4,2) id 0.3578
UN(4,3) id 0.4238
UN(4,4) id 0.5230
Fit Statistics
-2 Res Log Likelihood 385.1
AIC (smaller is better) 405.1
AICC (smaller is better) 406.1
BIC (smaller is better) 426.7
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
*RI attenders;
proc mixed data = pain1 method = reml noclprint noitprint noinfo;
where cs = "attender";
class tmt id trial;
model logpain = tmt/ notest;
repeated trial / subject = id type = cs;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Estimate
CS id 0.3003
Residual 0.09967
Fit Statistics
-2 Res Log Likelihood 151.3
AIC (smaller is better) 155.3
AICC (smaller is better) 155.4
BIC (smaller is better) 158.2
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 85.47 <.0001
*UN attenders;
proc mixed data = pain1 method = reml noclprint noitprint noinfo;
where cs = "attender";
class tmt id trial;
model logpain = tmt/ notest;
repeated trial / subject = id type = un;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Estimate
UN(1,1) id 0.3800
UN(2,1) id 0.3599
UN(2,2) id 0.4801
UN(3,1) id 0.3360
UN(3,2) id 0.3390
UN(3,3) id 0.3980
UN(4,1) id 0.2244
UN(4,2) id 0.2959
UN(4,3) id 0.2655
UN(4,4) id 0.3582
Fit Statistics
-2 Res Log Likelihood 134.8
AIC (smaller is better) 154.8
AICC (smaller is better) 156.9
BIC (smaller is better) 169.5
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 101.94 <.0001
*RI distracter;
proc mixed data = pain1 method = reml noclprint noitprint noinfo;
where cs = "distracter";
class tmt id trial;
model logpain = tmt/ notest;
repeated trial / subject = id type = cs;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Estimate
CS id 0.4630
Residual 0.2294
Fit Statistics
-2 Res Log Likelihood 240.5
AIC (smaller is better) 244.5
AICC (smaller is better) 244.6
BIC (smaller is better) 247.4
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 64.80 <.0001
*UN distracter;
proc mixed data = pain1 method = reml noclprint noitprint noinfo;
where cs = "distracter";
class tmt id trial;
model logpain = tmt/ notest;
repeated trial / subject = id type = un;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Estimate
UN(1,1) id 0.5874
UN(2,1) id 0.3621
UN(2,2) id 0.6100
UN(3,1) id 0.5769
UN(3,2) id 0.4978
UN(3,3) id 0.8828
UN(4,1) id 0.3323
UN(4,2) id 0.4283
UN(4,3) id 0.5878
UN(4,4) id 0.6941
Fit Statistics
-2 Res Log Likelihood 226.4
AIC (smaller is better) 246.4
AICC (smaller is better) 248.4
BIC (smaller is better) 261.0
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 78.89 <.0001
Table 8.9, page 284.
proc mixed data = pain1 method = reml covtest noclprint noitprint noinfo;
class cs id trial tmt;
model logpain = cs tmt/ notest;
repeated trial / subject = id type = cs group = cs;
run;
The Mixed Procedure
Covariance Parameter Estimates
Standard Z
Cov Parm Subject Group Estimate Error Value Pr Z
Variance id cs attender 0.1009 0.01540 6.55 <.0001
CS id cs attender 0.3034 0.08458 3.59 0.0003
Variance id cs distracter 0.2338 0.03513 6.66 <.0001
CS id cs distracter 0.4846 0.1389 3.49 0.0005
Fit Statistics
-2 Res Log Likelihood 395.2
AIC (smaller is better) 403.2
AICC (smaller is better) 403.4
BIC (smaller is better) 411.8
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
3 163.06 <.0001
Table 8.10, page 284.
proc mixed data = pain1 method = reml covtest noclprint noitprint noinfo;
class cs id trial tmt;
model logpain = cs tmt/ notest;
repeated trial / subject = id type = un group = cs;
run;
The Mixed Procedure
Covariance Parameter Estimates
Standard Z
Cov Parm Subject Group Estimate Error Value Pr Z
UN(1,1) id cs attender 0.3786 0.09606 3.94 <.0001
UN(2,1) id cs attender 0.3600 0.1010 3.57 0.0004
UN(2,2) id cs attender 0.4818 0.1260 3.82 <.0001
UN(3,1) id cs attender 0.3354 0.09244 3.63 0.0003
UN(3,2) id cs attender 0.3398 0.1022 3.32 0.0009
UN(3,3) id cs attender 0.3981 0.1029 3.87 <.0001
UN(4,1) id cs attender 0.2320 0.08128 2.85 0.0043
UN(4,2) id cs attender 0.3032 0.09910 3.06 0.0022
UN(4,3) id cs attender 0.2753 0.08880 3.10 0.0019
UN(4,4) id cs attender 0.3778 0.1043 3.62 0.0001
UN(1,1) id cs distracter 0.5882 0.1505 3.91 <.0001
UN(2,1) id cs distracter 0.3622 0.1260 2.88 0.0040
UN(2,2) id cs distracter 0.6082 0.1542 3.94 <.0001
UN(3,1) id cs distracter 0.5780 0.1673 3.46 0.0006
UN(3,2) id cs distracter 0.5002 0.1608 3.11 0.0019
UN(3,3) id cs distracter 0.8874 0.2290 3.88 <.0001
UN(4,1) id cs distracter 0.3808 0.1437 2.65 0.0080
UN(4,2) id cs distracter 0.4552 0.1513 3.01 0.0026
UN(4,3) id cs distracter 0.6564 0.1965 3.34 0.0008
UN(4,4) id cs distracter 0.8053 0.2135 3.77 <.0001
Fit Statistics
-2 Res Log Likelihood 364.8
AIC (smaller is better) 404.8
AICC (smaller is better) 408.6
BIC (smaller is better) 448.0
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
19 193.44 <.0001
Table 8.11, page 285.
*Attenders;
proc mixed data = pain2 method = reml covtest noclprint noitprint noinfo;
where cs = "attender";
class cs id trial ;
model logpain = cs tmt_a tmt_d tmt_n/ notest ;
repeated trial / subject = id type = un rcorr;
run;
The Mixed Procedure
Estimated R Correlation Matrix for id 1
Row Col1 Col2 Col3 Col4
1 1.0000 0.8425 0.8641 0.6083
2 0.8425 1.0000 0.7757 0.7136
3 0.8641 0.7757 1.0000 0.7033
4 0.6083 0.7136 0.7033 1.0000
Covariance Parameter Estimates
Standard Z
Cov Parm Subject Estimate Error Value Pr Z
UN(1,1) id 0.3800 0.09678 3.93 <.0001
UN(2,1) id 0.3599 0.1009 3.57 0.0004
UN(2,2) id 0.4801 0.1254 3.83 <.0001
UN(3,1) id 0.3360 0.09276 3.62 0.0003
UN(3,2) id 0.3390 0.1019 3.33 0.0009
UN(3,3) id 0.3980 0.1029 3.87 <.0001
UN(4,1) id 0.2244 0.07922 2.83 0.0046
UN(4,2) id 0.2959 0.09668 3.06 0.0022
UN(4,3) id 0.2655 0.08637 3.07 0.0021
UN(4,4) id 0.3582 0.09872 3.63 0.0001
Fit Statistics
-2 Res Log Likelihood 134.8
AIC (smaller is better) 154.8
AICC (smaller is better) 156.9
BIC (smaller is better) 169.5
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 101.94 <.0001
*Distracters;
proc mixed data = pain2 method = reml noinfo covtest noclprint noitprint;
where cs = "distracter";
class cs id trial ;
model logpain = cs tmt_a tmt_d tmt_n/ notest;
repeated trial / subject = id type = un rcorr;
run;
The Mixed Procedure
Estimated R Correlation Matrix for id 2
Row Col1 Col2 Col3 Col4
1 1.0000 0.6050 0.8011 0.5204
2 0.6050 1.0000 0.6784 0.6583
3 0.8011 0.6784 1.0000 0.7509
4 0.5204 0.6583 0.7509 1.0000
Covariance Parameter Estimates
Standard Z
Cov Parm Subject Estimate Error Value Pr Z
UN(1,1) id 0.5874 0.1503 3.91 <.0001
UN(2,1) id 0.3621 0.1263 2.87 0.0041
UN(2,2) id 0.6100 0.1550 3.93 <.0001
UN(3,1) id 0.5769 0.1669 3.46 0.0005
UN(3,2) id 0.4978 0.1609 3.09 0.0020
UN(3,3) id 0.8828 0.2277 3.88 <.0001
UN(4,1) id 0.3323 0.1330 2.50 0.0125
UN(4,2) id 0.4283 0.1412 3.03 0.0024
UN(4,3) id 0.5878 0.1809 3.25 0.0012
UN(4,4) id 0.6941 0.1822 3.81 <.0001
Fit Statistics
-2 Res Log Likelihood 226.4
AIC (smaller is better) 246.4
AICC (smaller is better) 248.4
BIC (smaller is better) 261.0
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 78.89 <.0001
Table 8.12, page 286.NOTE: SAS does not allow for random statement with groups in proc mixed, which is what would best fit the description of the bottom model in the book. However, the results were matched using a repeated statement in place of a random statement.
*TOP MODEL;
proc mixed data = pain2 method = reml noclprint noitprint noinfo;
class id trial cs;
model logpain = attend tmt_a tmt_d tmt_n attend*tmt_a attend*tmt_d attend*tmt_n/ notest solution;
repeated trial / subject = id type = un group=cs;
estimate 'DA' tmt_a 1 ;
estimate 'DD' tmt_d 1 ;
estimate 'DN' tmt_n 1 ;
estimate 'AA' attend*tmt_a 1 tmt_a 1 ;
estimate 'AD' attend*tmt_d 1 tmt_d 1 ;
estimate 'AN' attend*tmt_n 1 tmt_n 1 ;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Group Estimate
UN(1,1) id cs attender 0.3800
UN(2,1) id cs attender 0.3599
UN(2,2) id cs attender 0.4801
UN(3,1) id cs attender 0.3360
UN(3,2) id cs attender 0.3390
UN(3,3) id cs attender 0.3980
UN(4,1) id cs attender 0.2244
UN(4,2) id cs attender 0.2959
UN(4,3) id cs attender 0.2655
UN(4,4) id cs attender 0.3582
UN(1,1) id cs distracter 0.5874
UN(2,1) id cs distracter 0.3621
UN(2,2) id cs distracter 0.6100
UN(3,1) id cs distracter 0.5769
UN(3,2) id cs distracter 0.4978
UN(3,3) id cs distracter 0.8828
UN(4,1) id cs distracter 0.3323
UN(4,2) id cs distracter 0.4283
UN(4,3) id cs distracter 0.5878
UN(4,4) id cs distracter 0.6941
Fit Statistics
-2 Res Log Likelihood 361.2
AIC (smaller is better) 401.2
AICC (smaller is better) 405.1
BIC (smaller is better) 444.4
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
19 189.96 <.0001
Solution for Fixed Effects
Standard
Effect Estimate Error DF t Value Pr > |t|
Intercept 3.5252 0.1214 62 29.05 <.0001
attend -0.3483 0.1614 62 -2.16 0.0348
tmt_a -0.1413 0.1637 175 -0.86 0.3895
tmt_d 0.4777 0.1565 175 3.05 0.0026
tmt_n -0.3339 0.1691 175 -1.97 0.0499
attend*tmt_a 0.2451 0.2061 175 1.19 0.2359
attend*tmt_d -0.4042 0.2004 175 -2.02 0.0452
attend*tmt_n 0.3334 0.2111 175 1.58 0.1160
Estimates
Standard
Label Estimate Error DF t Value Pr > |t|
DA -0.1413 0.1637 175 -0.86 0.3895
DD 0.4777 0.1565 175 3.05 0.0026
DN -0.3339 0.1691 175 -1.97 0.0499
AA 0.1039 0.1251 175 0.83 0.4076
AD 0.07351 0.1251 175 0.59 0.5577
AN -0.00047 0.1263 175 -0.00 0.9970
*Bottom MODEL;
proc mixed data = pain2 method = reml noclprint noitprint noinfo;
class id trial cs;
model logpain = attend tmt_a tmt_d tmt_n attend*tmt_a attend*tmt_d attend*tmt_n/ notest solution;
repeated trial / subject = id type = cs group=cs;
estimate 'DA' tmt_a 1 ;
estimate 'DD' tmt_d 1 ;
estimate 'DN' tmt_n 1 ;
estimate 'AA' attend*tmt_a 1 tmt_a 1 ;
estimate 'AD' attend*tmt_d 1 tmt_d 1 ;
estimate 'AN' attend*tmt_n 1 tmt_n 1 ;
run;
The Mixed Procedure
Covariance Parameter Estimates
Cov Parm Subject Group Estimate
Variance id cs attender 0.09966
CS id cs attender 0.3004
Variance id cs distracter 0.2294
CS id cs distracter 0.4630
Fit Statistics
-2 Res Log Likelihood 391.7
AIC (smaller is better) 399.7
AICC (smaller is better) 399.9
BIC (smaller is better) 408.4
Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
3 159.40 <.0001
Solution for Fixed Effects
Standard
Effect Estimate Error DF t Value Pr > |t|
Intercept 3.4736 0.1305 62 26.62 <.0001
attend -0.3307 0.1659 62 -1.99 0.0507
tmt_a -0.2440 0.1727 175 -1.41 0.1596
tmt_d 0.4035 0.1648 175 2.45 0.0153
tmt_n -0.3388 0.1758 175 -1.93 0.0556
attend*tmt_a 0.3171 0.2071 175 1.53 0.1276
attend*tmt_d -0.3730 0.2006 175 -1.86 0.0646
attend*tmt_n 0.2652 0.2100 175 1.26 0.2083
Estimates
Standard
Label Estimate Error DF t Value Pr > |t|
DA -0.2440 0.1727 175 -1.41 0.1596
DD 0.4035 0.1648 175 2.45 0.0153
DN -0.3388 0.1758 175 -1.93 0.0556
AA 0.07308 0.1143 175 0.64 0.5234
AD 0.03046 0.1143 175 0.27 0.7902
AN -0.07359 0.1148 175 -0.64 0.5225
[http://www.ats.ucla.edu/stat/sas/footer.htm]