******************************************************************************
*****                   EXAMPLE 6 MODELS                          ******
******************************************************************************.

SORT CASES BY PairID TwinID Wave.
* Getting person gripp means to use as level-2 predictor.
COMPUTE PMgripp = MEAN(gripp1 TO gripp5).
EXECUTE.
* Getting pair gripp means to use as level-3 predictor.
AGGREGATE /OUTFILE=* MODE=ADDVARIABLES /PRESORTED /BREAK = PairID /FMgripp = MEAN(PMgripp).
VARIABLE LABELS PMgripp "PMgripp: Person Mean Gripp" FMgripp "FMgripp: Family Mean Gripp".

*** Age Variables ***.
* Centering age at time 1 at 85 to use at level-3.
COMPUTE BFage85 = agew1 - 85.
* Within-person centering age at level-1 (like PERSON MEAN CENTERING).
COMPUTE time = age - agew1.
 VARIABLE LABELS BFage85 "BFage85: Age at Time1 (0=85)"
                 time    "time: Time Since Entry (0= Age Wave 1)".

*** Grip Strength Variables ***.
* Centering family mean gripp at 9 to use at level-3.
COMPUTE BFgripp9 = FMgripp - 9.
* Centering person mean gripp at 9 to use at level-2.
COMPUTE BPgripp9 = PMgripp - 9.
COMPUTE WFgripp  = PMgripp - FMgripp.
* Centering time-varying gripp to use at level-1.
COMPUTE TVgripp9 = gripp - 9.
COMPUTE WPgripp  = gripp - PMgripp.
VARIABLE LABELS 
BFgripp9 "BFgripp9: Between-Family Mean Grip Strength in Pounds (0=9)"
BPgripp9 "BPgripp9: Between-Person Mean Grip Strength in Pounds (0=9)"
WFgripp  "WFgripp: Within-Family Deviation from Mean Grip Strength in Pounds"
TVgripp9 "TVgripp9: Time-Varying Grip Strength in Pounds (0=9)"
WPgripp  "WPgripp: Within-Person Deviation from Mean Grip Strength in Pounds".

* Selecting only complete cases.
SELECT IF (NMISS(agew1, age, FMgripp, PMgripp, gripp, info)=0).
EXECUTE.


*** Unconditional Growth Models .

TITLE "SPSS Model 1a: Empty Means, 2-Level Model for Information Test Outcome".

>Warning # 2003.  Command name: TITLE
>The title given exceeds 60 characters in length.  The first 60 characters will
>be used.


SPSS Model 1a: Empty Means, 2-Level Model for Information Te


MIXED info BY PairID TwinID
 /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  =
  /RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:25
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED info BY PairID TwinID
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED =
/RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).
Resources Processor Time 00:00:00.06
Elapsed Time 00:00:00.17


[DataSet1] 


Warnings
A(n) FIXED subcommand has no specifications and will therefore be ignored.
The covariance structure for random effect with only one level will be changed to Identity.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
Random Effects Intercept 1 Identity 1 PairID * TwinID
Residual     1  
Total 2   3  
a. Dependent Variable: Information Test (0-44) (546+547).

Information Criteriaa
-2 Restricted Log Likelihood 12147.440
Akaike's Information Criterion (AIC) 12151.440
Hurvich and Tsai's Criterion (AICC) 12151.447
Bozdogan's Criterion (CAIC) 12164.356
Schwarz's Bayesian Criterion (BIC) 12162.356
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Information Test (0-44) (546+547).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 582.839 2689.416 .000
a. Dependent Variable: Information Test (0-44) (546+547).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 25.462935 .490998 582.839 51.860 .000 24.498595 26.427276
a. Dependent Variable: Information Test (0-44) (546+547).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 26.668166 1.120233 23.806 .000 24.560504 28.956698
Intercept [subject = PairID * TwinID] Variance 130.522143 8.383683 15.569 .000 115.082667 148.032977
a. Dependent Variable: Information Test (0-44) (546+547).


TITLE "SPSS Model 1b: Empty Means, 3-Level Model for Information Test Outcome".

>Warning # 2003.  Command name: TITLE
>The title given exceeds 60 characters in length.  The first 60 characters will
>be used.


SPSS Model 1b: Empty Means, 3-Level Model for Information Te


MIXED info BY PairID TwinID
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  =
  /RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:26
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED info BY PairID TwinID
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED =
/RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).
Resources Processor Time 00:00:00.05
Elapsed Time 00:00:00.14


[DataSet1] 


Warnings
A(n) FIXED subcommand has no specifications and will therefore be ignored.
The covariance structure for random effect with only one level will be changed to Identity.
The covariance structure for random effect with only one level will be changed to Identity.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
Random Effects Intercept 1 Identity 1 PairID
Intercept 1 Identity 1 PairID * TwinID
Residual     1  
Total 3   4  
a. Dependent Variable: Information Test (0-44) (546+547).

Information Criteriaa
-2 Restricted Log Likelihood 12045.940
Akaike's Information Criterion (AIC) 12051.940
Hurvich and Tsai's Criterion (AICC) 12051.954
Bozdogan's Criterion (CAIC) 12071.313
Schwarz's Bayesian Criterion (BIC) 12068.313
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Information Test (0-44) (546+547).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 326.586 1787.757 .000
a. Dependent Variable: Information Test (0-44) (546+547).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 25.210180 .596241 326.586 42.282 .000 24.037222 26.383137
a. Dependent Variable: Information Test (0-44) (546+547).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 26.754972 1.126957 23.741 .000 24.634894 29.057504
Intercept [subject = PairID] Variance 83.734948 9.817701 8.529 .000 66.543499 105.367792
Intercept [subject = PairID * TwinID] Variance 47.335611 5.399656 8.766 .000 37.852096 59.195138
a. Dependent Variable: Information Test (0-44) (546+547).


TITLE "SPSS Age Model: Empty Means, 3-Level Model for Age Predictor".


SPSS Age Model: Empty Means, 3-Level Model for Age Predictor


MIXED age BY PairID TwinID
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  =
  /RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:27
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED age BY PairID TwinID
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED =
/RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).
Resources Processor Time 00:00:00.06
Elapsed Time 00:00:00.14


[DataSet1] 


Warnings
A(n) FIXED subcommand has no specifications and will therefore be ignored.
The covariance structure for random effect with only one level will be changed to Identity.
The covariance structure for random effect with only one level will be changed to Identity.
The final Hessian matrix is not positive definite although all convergence criteria are satisfied. The MIXED procedure continues despite this warning. Validity of subsequent results cannot be ascertained.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
Random Effects Intercept 1 Identity 1 PairID
Intercept 1 Identity 1 PairID * TwinID
Residual     1  
Total 3   4  
a. Dependent Variable: Age at Wave.

Information Criteriaa
-2 Restricted Log Likelihood 8948.538
Akaike's Information Criterion (AIC) 8954.538
Hurvich and Tsai's Criterion (AICC) 8954.552
Bozdogan's Criterion (CAIC) 8973.911
Schwarz's Bayesian Criterion (BIC) 8970.911
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Age at Wave.


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 305.287 294109.449 .000
a. Dependent Variable: Age at Wave.

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 85.964762 .158513 305.287 542.319 .000 85.652844 86.276679
a. Dependent Variable: Age at Wave.


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 7.466047 .284196 26.271 .000 6.929304 8.044366
Intercept [subject = PairID] Variance 6.553371 .675250 9.705 .000 5.354984 8.019943
Intercept [subject = PairID * TwinID] Variance .000000b .000000 . . . .
a. Dependent Variable: Age at Wave.
b. This covariance parameter is redundant. The test statistic and confidence interval cannot be computed.


TITLE "SPSS Grip Model: Empty Means, 3-Level Model for Grip Strength Predictor".

>Warning # 2003.  Command name: TITLE
>The title given exceeds 60 characters in length.  The first 60 characters will
>be used.


SPSS Grip Model: Empty Means, 3-Level Model for Grip Strengt


MIXED gripp BY PairID TwinID
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  =
  /RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:27
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED gripp BY PairID TwinID
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED =
/RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).
Resources Processor Time 00:00:00.06
Elapsed Time 00:00:00.20


[DataSet1] 


Warnings
A(n) FIXED subcommand has no specifications and will therefore be ignored.
The covariance structure for random effect with only one level will be changed to Identity.
The covariance structure for random effect with only one level will be changed to Identity.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
Random Effects Intercept 1 Identity 1 PairID
Intercept 1 Identity 1 PairID * TwinID
Residual     1  
Total 3   4  
a. Dependent Variable: Maximum Grip Strength in Pounds (Gripb x 14.5).

Information Criteriaa
-2 Restricted Log Likelihood 7835.051
Akaike's Information Criterion (AIC) 7841.051
Hurvich and Tsai's Criterion (AICC) 7841.065
Bozdogan's Criterion (CAIC) 7860.424
Schwarz's Bayesian Criterion (BIC) 7857.424
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Maximum Grip Strength in Pounds (Gripb x 14.5).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 318.392 4042.048 .000
a. Dependent Variable: Maximum Grip Strength in Pounds (Gripb x 14.5).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 8.065990 .126869 318.392 63.577 .000 7.816382 8.315598
a. Dependent Variable: Maximum Grip Strength in Pounds (Gripb x 14.5).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 3.049563 .127155 23.983 .000 2.810255 3.309249
Intercept [subject = PairID] Variance 3.085846 .467364 6.603 .000 2.293275 4.152335
Intercept [subject = PairID * TwinID] Variance 2.552534 .343661 7.427 .000 1.960513 3.323329
a. Dependent Variable: Maximum Grip Strength in Pounds (Gripb x 14.5).


TITLE "SPSS Model 2a: Fixed Quadratic, Random Intercept for Pair and Twin".

>Warning # 2003.  Command name: TITLE
>The title given exceeds 60 characters in length.  The first 60 characters will
>be used.


SPSS Model 2a: Fixed Quadratic, Random Intercept for Pair an


MIXED info BY PairID TwinID WITH BFage85 time
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  = BFage85 time time*time
  /RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:28
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED info BY PairID TwinID WITH BFage85 time
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED = BFage85 time time*time
/RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT | SUBJECT(PairID*TwinID) COVTYPE(UN).
Resources Processor Time 00:00:00.09
Elapsed Time 00:00:00.19


[DataSet1] 


Warnings
The covariance structure for random effect with only one level will be changed to Identity.
The covariance structure for random effect with only one level will be changed to Identity.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
BFage85 1   1  
time 1   1  
time * time 1   1  
Random Effects Intercept 1 Identity 1 PairID
Intercept 1 Identity 1 PairID * TwinID
Residual     1  
Total 6   7  
a. Dependent Variable: Information Test (0-44) (546+547).

Information Criteriaa
-2 Restricted Log Likelihood 11878.045
Akaike's Information Criterion (AIC) 11884.045
Hurvich and Tsai's Criterion (AICC) 11884.059
Bozdogan's Criterion (CAIC) 11903.413
Schwarz's Bayesian Criterion (BIC) 11900.413
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Information Test (0-44) (546+547).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 378.000 1348.754 .000
BFage85 1 353.746 17.277 .000
time 1 1186.528 2.605 .107
time * time 1 1168.082 8.821 .003
a. Dependent Variable: Information Test (0-44) (546+547).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 25.101030 .683479 378.000 36.725 .000 23.757133 26.444927
BFage85 -.807369 .194241 353.746 -4.157 .000 -1.189380 -.425357
time -.235091 .145668 1186.528 -1.614 .107 -.520886 .050704
time * time -.055585 .018715 1168.082 -2.970 .003 -.092305 -.018866
a. Dependent Variable: Information Test (0-44) (546+547).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 22.772185 .960104 23.718 .000 20.966068 24.733891
Intercept [subject = PairID] Variance 79.536597 9.694705 8.204 .000 62.634580 100.999645
Intercept [subject = PairID * TwinID] Variance 52.413569 5.679777 9.228 .000 42.384162 64.816246
a. Dependent Variable: Information Test (0-44) (546+547).


TITLE "SPSS Model 2b: Add Random Linear Slope for Twin".


SPSS Model 2b: Add Random Linear Slope for Twin


MIXED info BY PairID TwinID WITH BFage85 time
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  = BFage85 time time*time
  /RANDOM = INTERCEPT      | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN).



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:28
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED info BY PairID TwinID WITH BFage85 time
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED = BFage85 time time*time
/RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN).
Resources Processor Time 00:00:00.23
Elapsed Time 00:00:00.37


[DataSet1] 


Warnings
The covariance structure for random effect with only one level will be changed to Identity.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
BFage85 1   1  
time 1   1  
time * time 1   1  
Random Effects Intercept 1 Identity 1 PairID
Intercept + timeb 2 Unstructured 3 PairID * TwinID
Residual     1  
Total 7   9  
a. Dependent Variable: Information Test (0-44) (546+547).
b. As of version 11.5, the syntax rules for the RANDOM subcommand have changed. Your command syntax may yield results that differ from those produced by prior versions. If you are using version 11 syntax, please consult the current syntax reference guide for more information.

Information Criteriaa
-2 Restricted Log Likelihood 11745.999
Akaike's Information Criterion (AIC) 11755.999
Hurvich and Tsai's Criterion (AICC) 11756.033
Bozdogan's Criterion (CAIC) 11788.278
Schwarz's Bayesian Criterion (BIC) 11783.278
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Information Test (0-44) (546+547).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 350.200 1454.952 .000
BFage85 1 347.266 14.651 .000
time 1 1168.269 1.197 .274
time * time 1 943.302 38.126 .000
a. Dependent Variable: Information Test (0-44) (546+547).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 25.277220 .662682 350.200 38.144 .000 23.973883 26.580556
BFage85 -.730776 .190920 347.266 -3.828 .000 -1.106281 -.355271
time -.145471 .132939 1168.269 -1.094 .274 -.406297 .115355
time * time -.102142 .016542 943.302 -6.175 .000 -.134605 -.069678
a. Dependent Variable: Information Test (0-44) (546+547).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 15.122755 .832471 18.166 .000 13.576078 16.845640
Intercept [subject = PairID] Variance 80.103664 9.410725 8.512 .000 63.628498 100.844703
Intercept + time [subject = PairID * TwinID] UN (1,1) 44.312130 5.257736 8.428 .000 35.117663 55.913883
UN (2,1) 1.622180 .790024 2.053 .040 .073762 3.170598
UN (2,2) 1.178441 .180563 6.526 .000 .872740 1.591221
a. Dependent Variable: Information Test (0-44) (546+547).


TITLE "SPSS Model 2c: Add Random Linear Slope for Pair".


SPSS Model 2c: Add Random Linear Slope for Pair


MIXED info BY PairID TwinID WITH BFage85 time
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  = BFage85 time time*time
  /RANDOM = INTERCEPT time | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN).



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:29
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED info BY PairID TwinID WITH BFage85 time
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED = BFage85 time time*time
/RANDOM = INTERCEPT time | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN).
Resources Processor Time 00:00:00.31
Elapsed Time 00:00:00.41


[DataSet1] 


Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
BFage85 1   1  
time 1   1  
time * time 1   1  
Random Effects Intercept + timeb 2 Unstructured 3 PairID
Intercept + timeb 2 Unstructured 3 PairID * TwinID
Residual     1  
Total 8   11  
a. Dependent Variable: Information Test (0-44) (546+547).
b. As of version 11.5, the syntax rules for the RANDOM subcommand have changed. Your command syntax may yield results that differ from those produced by prior versions. If you are using version 11 syntax, please consult the current syntax reference guide for more information.

Information Criteriaa
-2 Restricted Log Likelihood 11745.215
Akaike's Information Criterion (AIC) 11759.215
Hurvich and Tsai's Criterion (AICC) 11759.280
Bozdogan's Criterion (CAIC) 11804.406
Schwarz's Bayesian Criterion (BIC) 11797.406
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Information Test (0-44) (546+547).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 348.334 1447.025 .000
BFage85 1 347.757 15.189 .000
time 1 1039.578 1.150 .284
time * time 1 943.832 37.796 .000
a. Dependent Variable: Information Test (0-44) (546+547).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 25.255023 .663911 348.334 38.040 .000 23.949245 26.560801
BFage85 -.743871 .190867 347.757 -3.897 .000 -1.119270 -.368472
time -.142939 .133292 1039.578 -1.072 .284 -.404491 .118614
time * time -.101691 .016541 943.832 -6.148 .000 -.134152 -.069230
a. Dependent Variable: Information Test (0-44) (546+547).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 15.114553 .831108 18.186 .000 13.570319 16.834513
Intercept + time [subject = PairID] UN (1,1) 80.861008 9.503706 8.508 .000 64.223839 101.808031
UN (2,1) -.732982 .925788 -.792 .429 -2.547493 1.081529
UN (2,2) .064033 .169718 .377 .706 .000355 11.547199
Intercept + time [subject = PairID * TwinID] UN (1,1) 44.007543 5.221052 8.429 .000 34.877121 55.528204
UN (2,1) 1.957107 .882669 2.217 .027 .227108 3.687106
UN (2,2) 1.116481 .241617 4.621 .000 .730538 1.706317
a. Dependent Variable: Information Test (0-44) (546+547).



TITLE "SPSS Model 3a: Grip Strength at each level via PERSON MEAN CENTERING".

>Warning # 2003.  Command name: TITLE
>The title given exceeds 60 characters in length.  The first 60 characters will
>be used.


SPSS Model 3a: Grip Strength at each level via PERSON MEAN C


MIXED info BY PairID TwinID WITH BFage85 time WPgripp WFgripp BFgripp9
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  = BFage85 time time*time WPgripp WFgripp BFgripp9
  /RANDOM = INTERCEPT      | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN)
 /TEST = "Level-2 Contextual Effect" WFgripp  1 WPgripp -1
 /TEST = "Level-3 Contextual Effect" BFgripp9 1 WFgripp -1.



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:30
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED info BY PairID TwinID WITH BFage85 time WPgripp WFgripp BFgripp9
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED = BFage85 time time*time WPgripp WFgripp BFgripp9
/RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN)
/TEST = "Level-2 Contextual Effect" WFgripp 1 WPgripp -1
/TEST = "Level-3 Contextual Effect" BFgripp9 1 WFgripp -1.
Resources Processor Time 00:00:00.34
Elapsed Time 00:00:00.59


[DataSet1] 


Warnings
The covariance structure for random effect with only one level will be changed to Identity.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
BFage85 1   1  
time 1   1  
time * time 1   1  
WPgripp 1   1  
WFgripp 1   1  
BFgripp9 1   1  
Random Effects Intercept 1 Identity 1 PairID
Intercept + timeb 2 Unstructured 3 PairID * TwinID
Residual     1  
Total 10   12  
a. Dependent Variable: Information Test (0-44) (546+547).
b. As of version 11.5, the syntax rules for the RANDOM subcommand have changed. Your command syntax may yield results that differ from those produced by prior versions. If you are using version 11 syntax, please consult the current syntax reference guide for more information.

Information Criteriaa
-2 Restricted Log Likelihood 11677.918
Akaike's Information Criterion (AIC) 11687.918
Hurvich and Tsai's Criterion (AICC) 11687.953
Bozdogan's Criterion (CAIC) 11720.189
Schwarz's Bayesian Criterion (BIC) 11715.189
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Information Test (0-44) (546+547).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 354.131 1290.203 .000
BFage85 1 349.421 3.250 .072
time 1 1171.160 .407 .524
time * time 1 954.475 37.323 .000
WPgripp 1 1184.229 26.377 .000
WFgripp 1 281.035 16.495 .000
BFgripp9 1 338.146 37.628 .000
a. Dependent Variable: Information Test (0-44) (546+547).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 27.043171 .752885 354.131 35.919 .000 25.562483 28.523858
BFage85 -.346326 .192110 349.421 -1.803 .072 -.724163 .031512
time .088451 .138619 1171.160 .638 .524 -.183518 .360420
time * time -.101003 .016533 954.475 -6.109 .000 -.133448 -.068558
WPgripp .503122 .097963 1184.229 5.136 .000 .310922 .695322
WFgripp .914361 .225135 281.035 4.061 .000 .471197 1.357525
BFgripp9 1.511445 .246397 338.146 6.134 .000 1.026781 1.996109
a. Dependent Variable: Information Test (0-44) (546+547).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 15.312283 .841307 18.201 .000 13.749032 17.053274
Intercept [subject = PairID] Variance 71.390800 8.596084 8.305 .000 56.383242 90.392926
Intercept + time [subject = PairID * TwinID] UN (1,1) 41.900587 5.043537 8.308 .000 33.094949 53.049158
UN (2,1) 1.224166 .724708 1.689 .091 -.196235 2.644567
UN (2,2) .994539 .164741 6.037 .000 .718827 1.376004
a. Dependent Variable: Information Test (0-44) (546+547).


Custom Hypothesis Test 1 (Level-2 Contextual Effect)


Contrast Estimatesa,b
Contrast Estimate Std. Error df Test Value t Sig. 95% Confidence Interval
Lower Bound Upper Bound
L1 .411239 .241575 363.788 0 1.702 .090 -.063820 .886298
a. Level-2 Contextual Effect
b. Dependent Variable: Information Test (0-44) (546+547).


Custom Hypothesis Test 2 (Level-3 Contextual Effect)


Contrast Estimatesa,b
Contrast Estimate Std. Error df Test Value t Sig. 95% Confidence Interval
Lower Bound Upper Bound
L1 .597084 .327515 579.647 0 1.823 .069 -.046177 1.240345
a. Level-3 Contextual Effect
b. Dependent Variable: Information Test (0-44) (546+547).


TITLE "SPSS Model 3b: Grip Strength Convergence across levels via GRAND MEAN CENTERING".

>Warning # 2003.  Command name: TITLE
>The title given exceeds 60 characters in length.  The first 60 characters will
>be used.


SPSS Model 3b: Grip Strength Convergence across levels via G


MIXED info BY PairID TwinID WITH BFage85 time TVgripp9 BPgripp9 BFgripp9
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  = BFage85 time time*time TVgripp9 BPgripp9 BFgripp9
  /RANDOM = INTERCEPT      | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN)
  /TEST = "Level-2 Within-Family Effect" TVgripp9 1 BPgripp9 1
 /TEST = "Level-3 Between-Pair  Effect" TVgripp9 1 BPgripp9 1 BFgripp9 1.



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:30
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED info BY PairID TwinID WITH BFage85 time TVgripp9 BPgripp9 BFgripp9
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED = BFage85 time time*time TVgripp9 BPgripp9 BFgripp9
/RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN)
/TEST = "Level-2 Within-Family Effect" TVgripp9 1 BPgripp9 1
/TEST = "Level-3 Between-Pair Effect" TVgripp9 1 BPgripp9 1 BFgripp9 1.
Resources Processor Time 00:00:00.36
Elapsed Time 00:00:00.59


[DataSet1] 


Warnings
The covariance structure for random effect with only one level will be changed to Identity.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
BFage85 1   1  
time 1   1  
time * time 1   1  
TVgripp9 1   1  
BPgripp9 1   1  
BFgripp9 1   1  
Random Effects Intercept 1 Identity 1 PairID
Intercept + timeb 2 Unstructured 3 PairID * TwinID
Residual     1  
Total 10   12  
a. Dependent Variable: Information Test (0-44) (546+547).
b. As of version 11.5, the syntax rules for the RANDOM subcommand have changed. Your command syntax may yield results that differ from those produced by prior versions. If you are using version 11 syntax, please consult the current syntax reference guide for more information.

Information Criteriaa
-2 Restricted Log Likelihood 11677.918
Akaike's Information Criterion (AIC) 11687.918
Hurvich and Tsai's Criterion (AICC) 11687.953
Bozdogan's Criterion (CAIC) 11720.189
Schwarz's Bayesian Criterion (BIC) 11715.189
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Information Test (0-44) (546+547).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 354.131 1290.203 .000
BFage85 1 349.421 3.250 .072
time 1 1171.160 .407 .524
time * time 1 954.475 37.323 .000
TVgripp9 1 1184.229 26.377 .000
BPgripp9 1 363.788 2.898 .090
BFgripp9 1 579.647 3.324 .069
a. Dependent Variable: Information Test (0-44) (546+547).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 27.043171 .752885 354.131 35.919 .000 25.562483 28.523858
BFage85 -.346326 .192110 349.421 -1.803 .072 -.724163 .031512
time .088451 .138619 1171.160 .638 .524 -.183518 .360420
time * time -.101003 .016533 954.475 -6.109 .000 -.133448 -.068558
TVgripp9 .503122 .097963 1184.229 5.136 .000 .310922 .695322
BPgripp9 .411239 .241575 363.788 1.702 .090 -.063820 .886298
BFgripp9 .597084 .327515 579.647 1.823 .069 -.046177 1.240345
a. Dependent Variable: Information Test (0-44) (546+547).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 15.312283 .841307 18.201 .000 13.749032 17.053274
Intercept [subject = PairID] Variance 71.390800 8.596084 8.305 .000 56.383242 90.392926
Intercept + time [subject = PairID * TwinID] UN (1,1) 41.900587 5.043537 8.308 .000 33.094949 53.049158
UN (2,1) 1.224166 .724708 1.689 .091 -.196235 2.644567
UN (2,2) .994539 .164741 6.037 .000 .718827 1.376004
a. Dependent Variable: Information Test (0-44) (546+547).


Custom Hypothesis Test 1 (Level-2 Within-Family Effect)


Contrast Estimatesa,b
Contrast Estimate Std. Error df Test Value t Sig. 95% Confidence Interval
Lower Bound Upper Bound
L1 .914361 .225135 281.035 0 4.061 .000 .471197 1.357525
a. Level-2 Within-Family Effect
b. Dependent Variable: Information Test (0-44) (546+547).


Custom Hypothesis Test 2 (Level-3 Between-Pair  Effect)


Contrast Estimatesa,b
Contrast Estimate Std. Error df Test Value t Sig. 95% Confidence Interval
Lower Bound Upper Bound
L1 1.511445 .246397 338.146 0 6.134 .000 1.026781 1.996109
a. Level-3 Between-Pair Effect
b. Dependent Variable: Information Test (0-44) (546+547).


TITLE "SPSS Model 3c: Grip Strength at Level 1 and Level 2&3 via PERSON MEAN CENTERING".

>Warning # 2003.  Command name: TITLE
>The title given exceeds 60 characters in length.  The first 60 characters will
>be used.


SPSS Model 3c: Grip Strength at Level 1 and Level 2&3 via PE


MIXED info BY PairID TwinID WITH BFage85 time WPgripp BPgripp9
  /METHOD = REML
  /PRINT  = SOLUTION TESTCOV
  /FIXED  = BFage85 time time*time WPgripp BPgripp9
  /RANDOM = INTERCEPT      | SUBJECT(PairID) COVTYPE(UN)
  /RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN)
 /TEST = "Level-2&3 Contextual Effect" BPgripp9 1 WPgripp -1.



Mixed Model Analysis


Notes
Output Created 08-MAY-2013 15:41:31
Comments  
Input Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data File 1734
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax MIXED info BY PairID TwinID WITH BFage85 time WPgripp BPgripp9
/METHOD = REML
/PRINT = SOLUTION TESTCOV
/FIXED = BFage85 time time*time WPgripp BPgripp9
/RANDOM = INTERCEPT | SUBJECT(PairID) COVTYPE(UN)
/RANDOM = INTERCEPT time | SUBJECT(PairID*TwinID) COVTYPE(UN)
/TEST = "Level-2&3 Contextual Effect" BPgripp9 1 WPgripp -1.
Resources Processor Time 00:00:00.30
Elapsed Time 00:00:00.33


[DataSet1] 


Warnings
The covariance structure for random effect with only one level will be changed to Identity.

Model Dimensiona
  Number of Levels Covariance Structure Number of Parameters Subject Variables
Fixed Effects Intercept 1   1  
BFage85 1   1  
time 1   1  
time * time 1   1  
WPgripp 1   1  
BPgripp9 1   1  
Random Effects Intercept 1 Identity 1 PairID
Intercept + timeb 2 Unstructured 3 PairID * TwinID
Residual     1  
Total 9   11  
a. Dependent Variable: Information Test (0-44) (546+547).
b. As of version 11.5, the syntax rules for the RANDOM subcommand have changed. Your command syntax may yield results that differ from those produced by prior versions. If you are using version 11 syntax, please consult the current syntax reference guide for more information.

Information Criteriaa
-2 Restricted Log Likelihood 11680.840
Akaike's Information Criterion (AIC) 11690.840
Hurvich and Tsai's Criterion (AICC) 11690.875
Bozdogan's Criterion (CAIC) 11723.114
Schwarz's Bayesian Criterion (BIC) 11718.114
The information criteria are displayed in smaller-is-better forms.
a. Dependent Variable: Information Test (0-44) (546+547).


Fixed Effects


Type III Tests of Fixed Effectsa
Source Numerator df Denominator df F Sig.
Intercept 1 415.729 1481.177 .000
BFage85 1 366.292 5.204 .023
time 1 1171.502 .426 .514
time * time 1 954.701 37.402 .000
WPgripp 1 1184.691 26.811 .000
BPgripp9 1 555.894 48.781 .000
a. Dependent Variable: Information Test (0-44) (546+547).

Estimates of Fixed Effectsa
Parameter Estimate Std. Error df t Sig. 95% Confidence Interval
Lower Bound Upper Bound
Intercept 26.476715 .687956 415.729 38.486 .000 25.124410 27.829021
BFage85 -.427543 .187418 366.292 -2.281 .023 -.796094 -.058991
time .090499 .138612 1171.502 .653 .514 -.181456 .362454
time * time -.101100 .016531 954.701 -6.116 .000 -.133542 -.068659
WPgripp .507100 .097934 1184.691 5.178 .000 .314957 .699244
BPgripp9 1.184309 .169566 555.894 6.984 .000 .851241 1.517377
a. Dependent Variable: Information Test (0-44) (546+547).


Covariance Parameters


Estimates of Covariance Parametersa
Parameter Estimate Std. Error Wald Z Sig. 95% Confidence Interval
Lower Bound Upper Bound
Residual 15.308102 .840856 18.205 .000 13.745669 17.048133
Intercept [subject = PairID] Variance 71.963222 8.654370 8.315 .000 56.851802 91.091313
Intercept + time [subject = PairID * TwinID] UN (1,1) 41.978279 5.046666 8.318 .000 33.165949 53.132081
UN (2,1) 1.234526 .722003 1.710 .087 -.180574 2.649625
UN (2,2) .995269 .164739 6.041 .000 .719528 1.376682
a. Dependent Variable: Information Test (0-44) (546+547).


Custom Hypothesis Test  (Level-2&3 Contextual Effect)


Contrast Estimatesa,b
Contrast Estimate Std. Error df Test Value t Sig. 95% Confidence Interval
Lower Bound Upper Bound
L1 .677209 .192579 849.425 0 3.517 .000 .299222 1.055196
a. Level-2&3 Contextual Effect
b. Dependent Variable: Information Test (0-44) (546+547).