Chapter 11a: Descriptive Statistics for Level-1 Time-Varying Student Variables

The MEANS Procedure

Variable Label N Mean Std Dev Variance Minimum Maximum
close
WSclose
victim
WSvictim
close: Time-Varying Student-Teacher Closeness
WSclose: Within-Student Student-Teacher Closeness (0=SM)
victim: Time-Varying Student-Perceived Victimization
WSvictim: Within-Student Student-Perceived Victimization (0=SM)
1731
1731
1731
1731
4.2108512
-2.82206E-18
3.1257461
-8.97927E-19
0.9523782
0.5205383
0.9910959
0.4747601
0.9070242
0.2709602
0.9822711
0.2253972
1.0000000
-2.6606831
1.2870090
-2.3333333
6.5485309
2.9333661
6.0000000
2.1917272


Chapter 11a: Descriptive Statistics for Level-1 Time-Varying Class Variables

The MEANS Procedure

Variable Label N Mean Std Dev Variance Minimum Maximum
emosup
WCemosup
emosup: Time-Varying Class Emotional Support
WCemosup: Within-Class Emotional Support (0=CM)
99
99
4.9886742
8.522924E-17
0.7390299
0.4279964
0.5461652
0.1831809
2.9687500
-1.5208333
6.2500000
1.1979167


Chapter 11a: Descriptive Statistics for Level-2 Student Variables

The MEANS Procedure

Variable Label N Mean Std Dev Variance Minimum Maximum
SMclose
WCclose
SMvictim
WCvictim
girl
SMclose: Student Mean Closeness
WCclose: Within-Class Student-Teacher Closeness (0=CM)
SMvictim: Student Mean Victimization
WCvictim: Within-Class Student-Perceived Victimization (0=CM)
girl: Student is Boy=0 or Girl=1
597
597
597
597
597
4.2025760
9.372737E-17
3.1269656
-7.58745E-17
0.4974874
0.8047855
0.7302906
0.8739274
0.8320328
0.5004130
0.6476797
0.5333243
0.7637491
0.6922785
0.2504131
1.0000000
-2.7977396
1.6699832
-1.6327213
0
5.6483916
1.7954553
6.0000000
2.6644696
1.0000000


Chapter 11a: Descriptive Statistics for Level-3 Class Variables

The MEANS Procedure

Variable Label N Mean Std Dev Variance Minimum Maximum
CMclose
CMvictim
CMemosup
CMgirl
classsize
grade35
CMclose: Class Mean Student-Teacher Closeness
CMvictim3: Class Mean Student-Perceived Victimization
CMemosup: Class Mean Emotional Support
CMgirl: Percentage Girls in Class
classsize: # Students in Class
grade35: Grade 3=0, Grade5=1
33
33
33
33
33
33
4.2167017
3.1345595
4.9886742
0.5007654
23.0909091
0.6060606
0.3561585
0.2782486
0.6087253
0.1036019
2.6500429
0.4961977
0.1268489
0.0774223
0.3705465
0.0107333
7.0227273
0.2462121
3.3535929
2.4988554
3.6458333
0.2941176
19.0000000
0
4.9377603
3.6952319
5.8750000
0.7647059
29.0000000
1.0000000


Ch 11a: Empty Means, Two-Level Model Predicting Student-Teacher Closeness
Occasions Within Students*Classes

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effect ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 2
Columns in X 1
Columns in Z Per Subject 1
Subjects 597
Max Obs Per Subject 3

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4748.15981984  
1 2 4283.51806040 0.00000270
2 1 4283.51656564 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID*StudentID 0.4985 0.03787 13.16 <.0001
Residual   0.4140 0.01741 23.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4283.5
AIC (smaller is better) 4287.5
AICC (smaller is better) 4287.5
BIC (smaller is better) 4296.3

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 464.64 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4283.5 2 4287.5 4287.5 4290.9 4296.3 4298.3

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2050 0.03283 592 128.09 <.0001 0.05 4.1405 4.2695


Eq 11a.1: Empty Means, Three-Level Model Predicting Student-Teacher Closeness
Level-1 Occasions Within Level-2 Students Within Level-3 Classes

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 3
Columns in X 1
Columns in Z Per Subject 26
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4748.15981984  
1 2 4243.53453491 0.00000655
2 1 4243.53099795 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID 0.09005 0.03090 2.91 0.0018
UN(1,1) ClassID*StudentID 0.4137 0.03398 12.17 <.0001
Residual   0.4142 0.01743 23.77 <.0001

Fit Statistics
-2 Res Log Likelihood 4243.5
AIC (smaller is better) 4249.5
AICC (smaller is better) 4249.5
BIC (smaller is better) 4254.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 504.63 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4243.5 3 4249.5 4249.5 4251.0 4254.0 4257.0

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2146 0.06073 31 69.40 <.0001 0.05 4.0907 4.3385


Likelihood Ratio Test for FitEmpty2C vs. FitEmpty3C

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitEmpty2C 4283.5 2 4287.5 4296.3 . . .
FitEmpty3C 4243.5 3 4249.5 4254.0 39.9856 1 2.5585E-10


Ch 11a: Empty Means, Two-Level Model Predicting Student-Perceived Victimization
Occasions Within Students*Classes

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable victim
Covariance Structure Unstructured
Subject Effect ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 2
Columns in X 1
Columns in Z Per Subject 1
Subjects 597
Max Obs Per Subject 3

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4886.03765625  
1 2 4174.28351521 0.00000021
2 1 4174.28340963 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID*StudentID 0.6416 0.04446 14.43 <.0001
Residual   0.3440 0.01445 23.81 <.0001

Fit Statistics
-2 Res Log Likelihood 4174.3
AIC (smaller is better) 4178.3
AICC (smaller is better) 4178.3
BIC (smaller is better) 4187.1

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 711.75 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4174.3 2 4178.3 4178.3 4181.7 4187.1 4189.1

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 3.1268 0.03573 595 87.51 <.0001 0.05 3.0567 3.1970


Eq 11a.1: Empty Means, Three-Level Model Predicting Student-Perceived Victimization
Level-1 Occasions Within Level-2 Students Within Level-3 Classes

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable victim
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 3
Columns in X 1
Columns in Z Per Subject 26
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4886.03765625  
1 2 4168.33405314 0.00000174
2 1 4168.33318669 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID 0.03261 0.01891 1.72 0.0423
UN(1,1) ClassID*StudentID 0.6107 0.04389 13.92 <.0001
Residual   0.3439 0.01445 23.81 <.0001

Fit Statistics
-2 Res Log Likelihood 4168.3
AIC (smaller is better) 4174.3
AICC (smaller is better) 4174.3
BIC (smaller is better) 4178.8

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 717.70 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4168.3 3 4174.3 4174.3 4175.8 4178.8 4181.8

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 3.1307 0.04723 30.2 66.28 <.0001 0.05 3.0342 3.2271


Likelihood Ratio Test for FitEmpty2V vs. FitEmpty3V

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitEmpty2V 4174.3 2 4178.3 4187.1 . . .
FitEmpty3V 4168.3 3 4174.3 4178.8 5.95022 1 0.014715


Eq 11a.1: Empty Means, Levels 1 and 3 Model Predicting Class Emotional Support
Level-1 Ocasions Within Level-3 Classes

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable emosup
Covariance Structure Unstructured
Subject Effect ClassID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 2
Columns in X 1
Columns in Z Per Subject 1
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 3823.41046712  
1 2 2089.04505269 0.00001577
2 1 2089.03620296 0.00000003
3 1 2089.03618529 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID 0.3732 0.09421 3.96 <.0001
Residual   0.1788 0.006137 29.14 <.0001

Fit Statistics
-2 Res Log Likelihood 2089.0
AIC (smaller is better) 2093.0
AICC (smaller is better) 2093.0
BIC (smaller is better) 2096.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 1734.37 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
2089.0 2 2093.0 2093.0 2094.0 2096.0 2098.0

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.9883 0.1068 32 46.69 <.0001 0.05 4.7706 5.2059


Eq 11a.1: Empty Means, Levels 1 and 3 Model Predicting Class Emotional Support
Level-1 Ocasions Within Level-3 Classes

The Mixed Procedure

Model Information
Data Set WORK.CLASSPERWAVE11A
Dependent Variable emosup
Covariance Structure Unstructured
Subject Effect ClassID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 2
Columns in X 1
Columns in Z Per Subject 1
Subjects 33
Max Obs Per Subject 3

Number of Observations
Number of Observations Read 99
Number of Observations Used 99
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 223.43336227  
1 1 200.16390940 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID 0.2799 0.09397 2.98 0.0014
Residual   0.2720 0.04735 5.74 <.0001

Fit Statistics
-2 Res Log Likelihood 200.2
AIC (smaller is better) 204.2
AICC (smaller is better) 204.3
BIC (smaller is better) 207.2

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 23.27 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
200.2 2 204.2 204.3 205.2 207.2 209.2

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.9887 0.1060 32 47.08 <.0001 0.05 4.7728 5.2045


Eq 11a.3: Saturated Means, Unstructured Level-3 and Level-2 Variances
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 12
Columns in X 4
Columns in Z Per Subject 3
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4733.42380051  
1 2 4131.07199043 0.00004388
2 1 4131.05059173 0.00000006
3 1 4131.05056214 0.00000000

Convergence criteria met.

Estimated R Matrix for ClassID*StudentID
130231 5023101
Row Col1 Col2 Col3
1 0.7210 0.4052 0.3796
2 0.4052 0.7914 0.4990
3 0.3796 0.4990 0.8892

Estimated R Correlation Matrix for
ClassID*StudentID 130231 5023101
Row Col1 Col2 Col3
1 1.0000 0.5364 0.4741
2 0.5364 1.0000 0.5949
3 0.4741 0.5949 1.0000

Estimated G Matrix
Row Effect ClassID: Class
ID Variable		 wave:
Wave of
Study
(1-3)		 Col1 Col2 Col3
1 wave 130231 1 0.02971 0.04421 0.07075
2 wave 130231 2 0.04421 0.09659 0.1353
3 wave 130231 3 0.07075 0.1353 0.2024

Estimated G Correlation Matrix
Row Effect ClassID: Class
ID Variable		 wave:
Wave of
Study
(1-3)		 Col1 Col2 Col3
1 wave 130231 1 1.0000 0.8254 0.9124
2 wave 130231 2 0.8254 1.0000 0.9676
3 wave 130231 3 0.9124 0.9676 1.0000

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02971 0.01782 1.67 0.0477
UN(2,1) ClassID 0.04421 0.02175 2.03 0.0421
UN(2,2) ClassID 0.09659 0.03628 2.66 0.0039
UN(3,1) ClassID 0.07075 0.02922 2.42 0.0155
UN(3,2) ClassID 0.1353 0.04495 3.01 0.0026
UN(3,3) ClassID 0.2024 0.06424 3.15 0.0008
UN(1,1) ClassID*StudentID 0.7210 0.04383 16.45 <.0001
UN(2,1) ClassID*StudentID 0.4052 0.03683 11.00 <.0001
UN(2,2) ClassID*StudentID 0.7914 0.04783 16.55 <.0001
UN(3,1) ClassID*StudentID 0.3796 0.03844 9.88 <.0001
UN(3,2) ClassID*StudentID 0.4990 0.04197 11.89 <.0001
UN(3,3) ClassID*StudentID 0.8892 0.05420 16.41 <.0001

Fit Statistics
-2 Res Log Likelihood 4131.1
AIC (smaller is better) 4155.1
AICC (smaller is better) 4155.2
BIC (smaller is better) 4173.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
11 602.37 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4131.1 12 4155.1 4155.2 4161.1 4173.0 4185.0

Solution for Fixed Effects
Effect wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept   4.0848 0.08776 31.2 46.55 <.0001 0.05 3.9058 4.2637
wave 1 0.2605 0.06531 31.2 3.99 0.0004 0.05 0.1273 0.3937
wave 2 0.1237 0.04547 31.8 2.72 0.0105 0.05 0.03106 0.2164
wave 3 0 . . . . . . .

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
wave 2 30.8 15.92 7.96 0.0003 0.0016

Least Squares Means
Effect wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
wave 1 4.3453 0.04643 31.8 93.59 <.0001 0.05 4.2507 4.4399
wave 2 4.2085 0.06553 30.5 64.23 <.0001 0.05 4.0747 4.3422
wave 3 4.0848 0.08776 31.2 46.55 <.0001 0.05 3.9058 4.2637

Differences of Least Squares Means
Effect wave:
Wave of
Study
(1-3)		 wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
wave 1 2 0.1368 0.04892 30.4 2.80 0.0089 0.05 0.03696 0.2366
wave 1 3 0.2605 0.06531 31.2 3.99 0.0004 0.05 0.1273 0.3937
wave 2 3 0.1237 0.04547 31.8 2.72 0.0105 0.05 0.03106 0.2164


Ch 11a: Piecewise Means, Level-3 and Level-2 Random Intercepts
Three-Level Model Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 3
Columns in X 3
Columns in Z Per Subject 26
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4733.42380051  
1 2 4203.29759448 0.00000682
2 1 4203.29404102 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID 0.09033 0.03094 2.92 0.0018
UN(1,1) ClassID*StudentID 0.4190 0.03389 12.36 <.0001
Residual   0.3973 0.01673 23.75 <.0001

Fit Statistics
-2 Res Log Likelihood 4203.3
AIC (smaller is better) 4209.3
AICC (smaller is better) 4209.3
BIC (smaller is better) 4213.8

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 530.13 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4203.3 3 4209.3 4209.3 4210.8 4213.8 4216.8

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.3520 0.06456 39.5 67.41 <.0001 0.05 4.2214 4.4825
time1 -0.1430 0.03730 1137 -3.83 0.0001 0.05 -0.2162 -0.06983
w3 0.01849 0.06438 1136 0.29 0.7741 0.05 -0.1078 0.1448

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 1137 14.70 14.70 0.0001 0.0001
w3 1 1136 0.08 0.08 0.7740 0.7741


Ch 11a: Piecewise Means, Add Level-2 Random Time Slope
Three-Level Model Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 5
Columns in X 3
Columns in Z Per Subject 51
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4733.42380051  
1 3 4171.37607853 0.00015983
2 1 4171.29409857 0.00000047
3 1 4171.29386607 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.07016 0.02629 2.67 0.0038
UN(1,1) ClassID*StudentID 0.4018 0.04389 9.15 <.0001
UN(2,1) ClassID*StudentID -0.01124 0.02168 -0.52 0.6043
UN(2,2) ClassID*StudentID 0.07133 0.01708 4.18 <.0001
Residual   0.3260 0.01944 16.77 <.0001

Fit Statistics
-2 Res Log Likelihood 4171.3
AIC (smaller is better) 4181.3
AICC (smaller is better) 4181.3
BIC (smaller is better) 4188.8

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
4 562.13 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4171.3 5 4181.3 4181.3 4183.8 4188.8 4193.8

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.3521 0.05836 35.7 74.58 <.0001 0.05 4.2337 4.4705
time1 -0.1439 0.03559 1023 -4.04 <.0001 0.05 -0.2138 -0.07407
w3 0.02009 0.05846 582 0.34 0.7312 0.05 -0.09473 0.1349

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 1023 16.35 16.35 <.0001 <.0001
w3 1 582 0.12 0.12 0.7311 0.7312


Likelihood Ratio Test for FitPieceRI2RI3C vs. FitPieceRL2RI3C

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitPieceRI2RI3C 4203.3 3 4209.3 4213.8 . . .
FitPieceRL2RI3C 4171.3 5 4181.3 4188.8 32.0002 2 .000000113


Ch 11a: Piecewise Means, Add Level-3 Random Time Slope
Three-Level Model Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 3
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4733.42380051  
1 2 4133.86657352 0.00004450
2 1 4133.84443858 0.00000018
3 1 4133.84435146 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02692 0.01662 1.62 0.0527
UN(2,1) ClassID 0.02128 0.008872 2.40 0.0165
UN(2,2) ClassID 0.02275 0.008923 2.55 0.0054
UN(1,1) ClassID*StudentID 0.4060 0.04445 9.13 <.0001
UN(2,1) ClassID*StudentID -0.00561 0.02065 -0.27 0.7859
UN(2,2) ClassID*StudentID 0.04908 0.01639 2.99 0.0014
Residual   0.3266 0.01948 16.77 <.0001

Fit Statistics
-2 Res Log Likelihood 4133.8
AIC (smaller is better) 4147.8
AICC (smaller is better) 4147.9
BIC (smaller is better) 4158.3

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 599.58 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4133.8 7 4147.8 4147.9 4151.4 4158.3 4165.3

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.3461 0.04571 34.3 95.09 <.0001 0.05 4.2533 4.4390
time1 -0.1392 0.04387 97.7 -3.17 0.0020 0.05 -0.2262 -0.05209
w3 0.01792 0.05849 581 0.31 0.7594 0.05 -0.09696 0.1328

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 97.7 10.06 10.06 0.0015 0.0020
w3 1 581 0.09 0.09 0.7593 0.7594


Likelihood Ratio Test for FitPieceRL2RI3C vs. FitPieceRL2RL3C

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitPieceRL2RI3C 4171.3 5 4181.3 4188.8 . . .
FitPieceRL2RL3C 4133.8 7 4147.8 4158.3 37.4495 2 7.378E-9


Likelihood Ratio Test for FitPieceRL2RL3C vs. FitSatUNC

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitPieceRL2RL3C 4133.8 7 4147.8 4158.3 . . .
FitSatUNC 4131.1 12 4155.1 4173.0 2.79379 5 0.73174


Eq 11a.4: Unconditional Growth Model Predicting Student Closeness
Random Linear Time Slopes for Level-2 Students and Level-3 Classes

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 2
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4730.68360497  
1 2 4130.11991522 0.00004479
2 1 4130.09776763 0.00000018
3 1 4130.09768060 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02692 0.01662 1.62 0.0527
UN(2,1) ClassID 0.02127 0.008873 2.40 0.0165
UN(2,2) ClassID 0.02276 0.008924 2.55 0.0054
UN(1,1) ClassID*StudentID 0.4064 0.04444 9.15 <.0001
UN(2,1) ClassID*StudentID -0.00583 0.02064 -0.28 0.7777
UN(2,2) ClassID*StudentID 0.04929 0.01637 3.01 0.0013
Residual   0.3261 0.01943 16.79 <.0001

Fit Statistics
-2 Res Log Likelihood 4130.1
AIC (smaller is better) 4144.1
AICC (smaller is better) 4144.2
BIC (smaller is better) 4154.6

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 600.59 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4130.1 7 4144.1 4144.2 4147.6 4154.6 4161.6

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.3431 0.04461 31.1 97.35 <.0001 0.05 4.2521 4.4340
time1 -0.1302 0.03270 31.2 -3.98 0.0004 0.05 -0.1969 -0.06351

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 31.2 15.85 15.85 <.0001 0.0004


Total R2 (% Reduction) for PredEmpty3C vs. PredUncC

Name PredCorr TotalR2 TotalR2Diff
PredEmpty3C . . .
PredUncC 0.11441 0.013090 .


Ch 11a: Add Class Grade and Class Size
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 6
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4710.72689733  
1 2 4144.30288046 0.00004022
2 1 4144.28266741 0.00000012
3 1 4144.28260782 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02311 0.01642 1.41 0.0796
UN(2,1) ClassID 0.01683 0.008611 1.95 0.0507
UN(2,2) ClassID 0.02154 0.008835 2.44 0.0074
UN(1,1) ClassID*StudentID 0.4067 0.04449 9.14 <.0001
UN(2,1) ClassID*StudentID -0.00583 0.02065 -0.28 0.7776
UN(2,2) ClassID*StudentID 0.04916 0.01637 3.00 0.0013
Residual   0.3262 0.01944 16.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4144.3
AIC (smaller is better) 4158.3
AICC (smaller is better) 4158.3
BIC (smaller is better) 4168.8

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 566.44 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4144.3 7 4158.3 4158.3 4161.8 4168.8 4175.8

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.3200 0.07074 29.7 61.07 <.0001 0.05 4.1754 4.4645
time1 -0.1490 0.05230 30.7 -2.85 0.0078 0.05 -0.2557 -0.04231
grade35 0.04563 0.09135 29 0.50 0.6212 0.05 -0.1412 0.2325
time1*grade35 0.03435 0.06767 30.1 0.51 0.6154 0.05 -0.1038 0.1725
size23 -0.03263 0.01703 28.7 -1.92 0.0653 0.05 -0.06748 0.002212
time1*size23 -0.02237 0.01259 29.5 -1.78 0.0859 0.05 -0.04810 0.003360

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 30.7 8.12 8.12 0.0044 0.0078
grade35 1 29 0.25 0.25 0.6174 0.6212
time1*grade35 1 30.1 0.26 0.26 0.6117 0.6154
size23 1 28.7 3.67 3.67 0.0553 0.0653
time1*size23 1 29.5 3.16 3.16 0.0756 0.0859


Ch 11a: Add Class Size Only
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 4
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4709.00004181  
1 2 4138.20196309 0.00002692
2 1 4138.18868472 0.00000004
3 1 4138.18866372 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02190 0.01567 1.40 0.0812
UN(2,1) ClassID 0.01703 0.008284 2.06 0.0399
UN(2,2) ClassID 0.02076 0.008490 2.45 0.0072
UN(1,1) ClassID*StudentID 0.4065 0.04446 9.14 <.0001
UN(2,1) ClassID*StudentID -0.00580 0.02065 -0.28 0.7789
UN(2,2) ClassID*StudentID 0.04919 0.01637 3.00 0.0013
Residual   0.3262 0.01944 16.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4138.2
AIC (smaller is better) 4152.2
AICC (smaller is better) 4152.3
BIC (smaller is better) 4162.7

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 570.81 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4138.2 7 4152.2 4152.3 4155.7 4162.7 4169.7

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.3476 0.04303 30.4 101.04 <.0001 0.05 4.2598 4.4355
time1 -0.1281 0.03183 30.9 -4.02 0.0003 0.05 -0.1931 -0.06319
size23 -0.03087 0.01646 30.3 -1.88 0.0704 0.05 -0.06448 0.002738
time1*size23 -0.02095 0.01215 30.6 -1.72 0.0947 0.05 -0.04574 0.003841

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 30.9 16.20 16.20 <.0001 0.0003
size23 1 30.3 3.52 3.52 0.0608 0.0704
time1*size23 1 30.6 2.97 2.97 0.0846 0.0947


PsuedoR2 (% Reduction) for CovUncC vs. CovSizeC

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
CovUncC UN(1,1) ClassID 0.02692 0.01662 1.62 0.0527 .
CovUncC UN(2,2) ClassID 0.02276 0.008924 2.55 0.0054 .
CovUncC UN(1,1) ClassID*StudentID 0.4064 0.04444 9.15 <.0001 .
CovUncC UN(2,2) ClassID*StudentID 0.04929 0.01637 3.01 0.0013 .
CovUncC Residual   0.3261 0.01943 16.79 <.0001 .
CovSizeC UN(1,1) ClassID 0.02190 0.01567 1.40 0.0812 0.18666
CovSizeC UN(2,2) ClassID 0.02076 0.008490 2.45 0.0072 0.08797
CovSizeC UN(1,1) ClassID*StudentID 0.4065 0.04446 9.14 <.0001 -0.00024
CovSizeC UN(2,2) ClassID*StudentID 0.04919 0.01637 3.00 0.0013 0.00211
CovSizeC Residual   0.3262 0.01944 16.78 <.0001 -0.00026


Eq 11a.5: Add Student and Class Gender
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 8
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4675.04678967  
1 2 4116.04543769 0.00262996
2 2 4115.04287203 0.00029274
3 1 4114.89048378 0.00000852
4 1 4114.88635668 0.00000001

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02260 0.01565 1.44 0.0744
UN(2,1) ClassID 0.02276 0.007483 3.04 0.0023
UN(2,2) ClassID 0.01259 0.006540 1.93 0.0271
UN(1,1) ClassID*StudentID 0.3907 0.04345 8.99 <.0001
UN(2,1) ClassID*StudentID -0.00519 0.02042 -0.25 0.7992
UN(2,2) ClassID*StudentID 0.04893 0.01633 3.00 0.0014
Residual   0.3261 0.01942 16.79 <.0001

Fit Statistics
-2 Res Log Likelihood 4114.9
AIC (smaller is better) 4128.9
AICC (smaller is better) 4129.0
BIC (smaller is better) 4139.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 560.16 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4114.9 7 4128.9 4129.0 4132.4 4139.4 4146.4

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2208 0.05497 78.9 76.78 <.0001 0.05 4.1114 4.3303
time1 -0.1243 0.03376 65.1 -3.68 0.0005 0.05 -0.1917 -0.05686
size23 -0.02857 0.01647 30.1 -1.73 0.0931 0.05 -0.06220 0.005066
time1*size23 -0.02356 0.01056 29.1 -2.23 0.0336 0.05 -0.04515 -0.00196
girl 0.2534 0.06882 556 3.68 0.0003 0.05 0.1182 0.3886
time1*girl -0.00629 0.03936 551 -0.16 0.8730 0.05 -0.08361 0.07102
CMgirl50 0.4998 0.4380 34.9 1.14 0.2616 0.05 -0.3895 1.3891
time1*CMgirl50 -0.9240 0.2811 33.7 -3.29 0.0024 0.05 -1.4954 -0.3526

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 65.1 13.55 13.55 0.0002 0.0005
size23 1 30.1 3.01 3.01 0.0828 0.0931
time1*size23 1 29.1 4.98 4.98 0.0257 0.0336
girl 1 556 13.56 13.56 0.0002 0.0003
time1*girl 1 551 0.03 0.03 0.8729 0.8730
CMgirl50 1 34.9 1.30 1.30 0.2538 0.2616
time1*CMgirl50 1 33.7 10.81 10.81 0.0010 0.0024

Estimates
Label Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Contextual Gender Effect at Wave 1 0.4998 0.4380 34.9 1.14 0.2616 0.05 -0.3895 1.3891
Contextual Gender Effect at Wave 2 -0.4242 0.5833 31.5 -0.73 0.4724 0.05 -1.6130 0.7646
Contextual Gender Effect at Wave 3 -1.3482 0.8041 31.3 -1.68 0.1036 0.05 -2.9877 0.2913

Contrasts
Label Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
Multivariate Test of Level-2 Within-Class Gender Effects 2 552 16.56 8.28 0.0003 0.0003
Multivariate Test of Level-3 Contextual Class Gender Effects 2 36.2 15.46 7.73 0.0004 0.0016


PsuedoR2 (% Reduction) for CovSizeC vs. CovGirlC

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
CovSizeC UN(1,1) ClassID 0.02190 0.01567 1.40 0.0812 .
CovSizeC UN(2,2) ClassID 0.02076 0.008490 2.45 0.0072 .
CovSizeC UN(1,1) ClassID*StudentID 0.4065 0.04446 9.14 <.0001 .
CovSizeC UN(2,2) ClassID*StudentID 0.04919 0.01637 3.00 0.0013 .
CovSizeC Residual   0.3262 0.01944 16.78 <.0001 .
CovGirlC UN(1,1) ClassID 0.02260 0.01565 1.44 0.0744 -0.03226
CovGirlC UN(2,2) ClassID 0.01259 0.006540 1.93 0.0271 0.39342
CovGirlC UN(1,1) ClassID*StudentID 0.3907 0.04345 8.99 <.0001 0.03901
CovGirlC UN(2,2) ClassID*StudentID 0.04893 0.01633 3.00 0.0014 0.00514
CovGirlC Residual   0.3261 0.01942 16.79 <.0001 0.00033


Ch 11a: Test if Level-3 Random Time Slope Variance is still needed
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 5
Columns in X 8
Columns in Z Per Subject 51
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4675.04678967  
1 3 4140.18872635 0.00008051
2 1 4140.14882834 0.00000007
3 1 4140.14879604 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.06225 0.02470 2.52 0.0059
UN(1,1) ClassID*StudentID 0.3840 0.04276 8.98 <.0001
UN(2,1) ClassID*StudentID -0.00476 0.02083 -0.23 0.8193
UN(2,2) ClassID*StudentID 0.06012 0.01659 3.62 0.0001
Residual   0.3261 0.01943 16.79 <.0001

Fit Statistics
-2 Res Log Likelihood 4140.1
AIC (smaller is better) 4150.1
AICC (smaller is better) 4150.2
BIC (smaller is better) 4157.6

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
4 534.90 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4140.1 5 4150.1 4150.2 4152.7 4157.6 4162.6

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2239 0.06495 60.2 65.04 <.0001 0.05 4.0940 4.3538
time1 -0.1261 0.02820 574 -4.47 <.0001 0.05 -0.1815 -0.07071
size23 -0.02807 0.02119 32.4 -1.32 0.1945 0.05 -0.07121 0.01507
time1*size23 -0.02310 0.007611 571 -3.03 0.0025 0.05 -0.03805 -0.00815
girl 0.2530 0.06851 566 3.69 0.0002 0.05 0.1185 0.3876
time1*girl -0.00529 0.04037 578 -0.13 0.8958 0.05 -0.08457 0.07399
CMgirl50 0.5409 0.5556 35.5 0.97 0.3369 0.05 -0.5864 1.6681
time1*CMgirl50 -0.9833 0.2095 579 -4.69 <.0001 0.05 -1.3949 -0.5718

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 574 19.99 19.99 <.0001 <.0001
size23 1 32.4 1.76 1.76 0.1852 0.1945
time1*size23 1 571 9.21 9.21 0.0024 0.0025
girl 1 566 13.64 13.64 0.0002 0.0002
time1*girl 1 578 0.02 0.02 0.8958 0.8958
CMgirl50 1 35.5 0.95 0.95 0.3303 0.3369
time1*CMgirl50 1 579 22.02 22.02 <.0001 <.0001


Likelihood Ratio Test for NoRandTime3C vs. FitGirlC

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
NoRandTime3C 4140.1 5 4150.1 4157.6 . . .
FitGirlC 4114.9 7 4128.9 4139.4 25.2624 2 .000003268


Ch 11a: Add Random Gender Effect across Classes
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 8
Columns in Z Per Subject 53
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4675.04678967  
1 2 4110.54718093 0.01832303
2 2 4107.60059190 0.00076918
3 2 4107.22067049 0.00000493
4 1 4107.21834100 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.06875 0.03667 1.87 0.0304
UN(2,1) ClassID 0.02501 0.01149 2.18 0.0295
UN(2,2) ClassID 0.01249 0.006523 1.92 0.0277
UN(3,1) ClassID -0.04671 0.03183 -1.47 0.1422
UN(3,2) ClassID -0.00312 0.01093 -0.29 0.7753
UN(3,3) ClassID 0.01260 0.03311 0.38 0.3518
UN(1,1) ClassID*StudentID 0.3887 0.04400 8.83 <.0001
UN(2,1) ClassID*StudentID -0.00601 0.02038 -0.30 0.7680
UN(2,2) ClassID*StudentID 0.04893 0.01631 3.00 0.0014
Residual   0.3258 0.01940 16.80 <.0001

Fit Statistics
-2 Res Log Likelihood 4107.2
AIC (smaller is better) 4127.2
AICC (smaller is better) 4127.3
BIC (smaller is better) 4142.2

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 567.83 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4107.2 10 4127.2 4127.3 4132.3 4142.2 4152.2

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2211 0.06658 31.1 63.39 <.0001 0.05 4.0853 4.3568
time1 -0.1243 0.03370 65 -3.69 0.0005 0.05 -0.1916 -0.05700
size23 -0.01794 0.01517 33.6 -1.18 0.2452 0.05 -0.04878 0.01290
time1*size23 -0.02314 0.01052 29.1 -2.20 0.0359 0.05 -0.04465 -0.00163
girl 0.2502 0.07133 49 3.51 0.0010 0.05 0.1068 0.3935
time1*girl -0.00562 0.03934 552 -0.14 0.8865 0.05 -0.08288 0.07165
CMgirl50 0.2838 0.3936 34 0.72 0.4759 0.05 -0.5162 1.0837
time1*CMgirl50 -0.9309 0.2800 33.6 -3.32 0.0021 0.05 -1.5002 -0.3616

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 65 13.61 13.61 0.0002 0.0005
size23 1 33.6 1.40 1.40 0.2369 0.2452
time1*size23 1 29.1 4.84 4.84 0.0278 0.0359
girl 1 49 12.30 12.30 0.0005 0.0010
time1*girl 1 552 0.02 0.02 0.8865 0.8865
CMgirl50 1 34 0.52 0.52 0.4710 0.4759
time1*CMgirl50 1 33.6 11.05 11.05 0.0009 0.0021


Likelihood Ratio Test for FitGirlC vs. FitRandGirlC

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitGirlC 4114.9 7 4128.9 4139.4 . . .
FitRandGirlC 4107.2 10 4127.2 4142.2 7.66802 3 0.053395


Ch 11a: Saturated Means, Unstructured Model Predicting Class Emotional Support
Using Only Level-1 Ocasions Within Level-3 Classes

The Mixed Procedure

Model Information
Data Set WORK.CLASSPERWAVE11A
Dependent Variable emosup
Covariance Structure Unstructured
Subject Effect ClassID
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 6
Columns in X 4
Columns in Z 0
Subjects 33
Max Obs Per Subject 3

Number of Observations
Number of Observations Read 99
Number of Observations Used 99
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 225.91639244  
1 1 199.19197792 0.00000000

Convergence criteria met.

Estimated R Matrix for ClassID 130231
Row Col1 Col2 Col3
1 0.5961 0.3551 0.2186
2 0.3551 0.6792 0.2654
3 0.2186 0.2654 0.3813

Estimated R Correlation Matrix for
ClassID 130231
Row Col1 Col2 Col3
1 1.0000 0.5581 0.4586
2 0.5581 1.0000 0.5216
3 0.4586 0.5216 1.0000

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.5961 0.1490 4.00 <.0001
UN(2,1) ClassID 0.3551 0.1288 2.76 0.0058
UN(2,2) ClassID 0.6792 0.1698 4.00 <.0001
UN(3,1) ClassID 0.2186 0.09272 2.36 0.0184
UN(3,2) ClassID 0.2654 0.1015 2.62 0.0089
UN(3,3) ClassID 0.3813 0.09532 4.00 <.0001

Fit Statistics
-2 Res Log Likelihood 199.2
AIC (smaller is better) 211.2
AICC (smaller is better) 212.1
BIC (smaller is better) 220.2

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
5 26.72 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
199.2 6 211.2 212.1 214.2 220.2 226.2

Solution for Fixed Effects
Effect wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept   5.0900 0.1075 32 47.35 <.0001 0.05 4.8710 5.3089
wave 1 -0.1609 0.1279 32 -1.26 0.2177 0.05 -0.4215 0.09972
wave 2 -0.1430 0.1267 32 -1.13 0.2674 0.05 -0.4011 0.1151
wave 3 0 . . . . . . .

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
wave 2 32 1.95 0.98 0.3772 0.3881

Least Squares Means
Effect wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
wave 1 4.9291 0.1344 32 36.67 <.0001 0.05 4.6553 5.2029
wave 2 4.9470 0.1435 32 34.48 <.0001 0.05 4.6547 5.2392
wave 3 5.0900 0.1075 32 47.35 <.0001 0.05 4.8710 5.3089

Differences of Least Squares Means
Effect wave:
Wave of
Study
(1-3)		 wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
wave 1 2 -0.01788 0.1309 32 -0.14 0.8922 0.05 -0.2844 0.2487
wave 1 3 -0.1609 0.1279 32 -1.26 0.2177 0.05 -0.4215 0.09972
wave 2 3 -0.1430 0.1267 32 -1.13 0.2674 0.05 -0.4011 0.1151


Ch 11a: Saturated Means, Random Intercept Model Predicting Class Emotional Support
Using Only Level-1 Ocasions Within Level-3 Classes

The Mixed Procedure

Model Information
Data Set WORK.CLASSPERWAVE11A
Dependent Variable emosup
Covariance Structure Unstructured
Subject Effect ClassID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 2
Columns in X 4
Columns in Z Per Subject 1
Subjects 33
Max Obs Per Subject 3

Number of Observations
Number of Observations Read 99
Number of Observations Used 99
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 225.91639244  
1 1 203.09994524 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID 0.2797 0.09402 2.98 0.0015
Residual   0.2725 0.04817 5.66 <.0001

Fit Statistics
-2 Res Log Likelihood 203.1
AIC (smaller is better) 207.1
AICC (smaller is better) 207.2
BIC (smaller is better) 210.1

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
1 22.82 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
203.1 2 207.1 207.2 208.1 210.1 212.1

Solution for Fixed Effects
Effect wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept   5.0900 0.1294 63.4 39.35 <.0001 0.05 4.8315 5.3484
wave 1 -0.1609 0.1285 64 -1.25 0.2152 0.05 -0.4176 0.09585
wave 2 -0.1430 0.1285 64 -1.11 0.2700 0.05 -0.3997 0.1137
wave 3 0 . . . . . . .

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
wave 2 64 1.88 0.94 0.3900 0.3954


Likelihood Ratio Test for FitSatRIE vs. FitSatUNE

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitSatRIE 203.1 2 207.1 210.1 . . .
FitSatUNE 199.2 6 211.2 220.2 3.90797 4 0.41860


Ch 11a: Saturated Means, Random Time Slope Model Predicting Class Emotional Support
Using Only Level-1 Ocasions Within Level-3 Classes

The Mixed Procedure

Model Information
Data Set WORK.CLASSPERWAVE11A
Dependent Variable emosup
Covariance Structure Unstructured
Subject Effect ClassID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 4
Columns in X 4
Columns in Z Per Subject 2
Subjects 33
Max Obs Per Subject 3

Number of Observations
Number of Observations Read 99
Number of Observations Used 99
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 225.91639244  
1 4 201.42622831 0.00143306
2 1 201.40861560 0.00000174
3 1 201.40859384 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.3823 0.1385 2.76 0.0029
UN(2,1) ClassID -0.05121 0.03990 -1.28 0.1993
UN(2,2) ClassID 0 . . .
Residual   0.2725 0.04817 5.66 <.0001

Fit Statistics
-2 Res Log Likelihood 201.4
AIC (smaller is better) 207.4
AICC (smaller is better) 207.7
BIC (smaller is better) 211.9

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 24.51 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
201.4 3 207.4 207.7 208.9 211.9 214.9

Solution for Fixed Effects
Effect wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept   5.0900 0.1168 39 43.59 <.0001 0.05 4.8538 5.3261
wave 1 -0.1609 0.1285 64 -1.25 0.2152 0.05 -0.4176 0.09585
wave 2 -0.1430 0.1285 64 -1.11 0.2700 0.05 -0.3997 0.1137
wave 3 0 . . . . . . .

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
wave 2 64 1.88 0.94 0.3900 0.3954


Eq 11a.6: Add Time-Varying and Class Emotional Support
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 12
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4649.43284302  
1 3 4121.92642470 0.01098370
2 2 4119.69231024 0.00094139
3 2 4119.25606683 0.00001783
4 1 4119.24731191 0.00000003
5 1 4119.24729902 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02660 0.01738 1.53 0.0630
UN(2,1) ClassID 0.02041 0.006552 3.12 0.0018
UN(2,2) ClassID 0.004074 0.004628 0.88 0.1894
UN(1,1) ClassID*StudentID 0.3889 0.04344 8.95 <.0001
UN(2,1) ClassID*StudentID -0.00346 0.02038 -0.17 0.8652
UN(2,2) ClassID*StudentID 0.04739 0.01632 2.90 0.0018
Residual   0.3279 0.01956 16.76 <.0001

Fit Statistics
-2 Res Log Likelihood 4119.2
AIC (smaller is better) 4133.2
AICC (smaller is better) 4133.3
BIC (smaller is better) 4143.7

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 530.19 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4119.2 7 4133.2 4133.3 4136.8 4143.7 4150.7

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2239 0.05614 69.8 75.23 <.0001 0.05 4.1119 4.3359
time1 -0.1309 0.02975 76.4 -4.40 <.0001 0.05 -0.1901 -0.07166
size23 -0.02523 0.01801 27 -1.40 0.1728 0.05 -0.06219 0.01173
time1*size23 -0.01181 0.009057 24.1 -1.30 0.2044 0.05 -0.03050 0.006872
girl 0.2536 0.06879 557 3.69 0.0002 0.05 0.1184 0.3887
time1*girl -0.00736 0.03929 553 -0.19 0.8516 0.05 -0.08453 0.06982
CMgirl50 0.5103 0.4569 32.6 1.12 0.2722 0.05 -0.4197 1.4403
time1*CMgirl50 -1.0835 0.2347 31.3 -4.62 <.0001 0.05 -1.5620 -0.6051
emosup5 -0.02358 0.06140 120 -0.38 0.7017 0.05 -0.1451 0.09799
time1*emosup5 0.05030 0.05510 126 0.91 0.3630 0.05 -0.05873 0.1593
CMemosup5 0.05631 0.1031 64.1 0.55 0.5870 0.05 -0.1497 0.2623
time1*CMemosup5 0.1091 0.06483 78.6 1.68 0.0965 0.05 -0.01998 0.2381

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 76.4 19.36 19.36 <.0001 <.0001
size23 1 27 1.96 1.96 0.1614 0.1728
time1*size23 1 24.1 1.70 1.70 0.1921 0.2044
girl 1 557 13.59 13.59 0.0002 0.0002
time1*girl 1 553 0.04 0.04 0.8515 0.8516
CMgirl50 1 32.6 1.25 1.25 0.2641 0.2722
time1*CMgirl50 1 31.3 21.32 21.32 <.0001 <.0001
emosup5 1 120 0.15 0.15 0.7010 0.7017
time1*emosup5 1 126 0.83 0.83 0.3613 0.3630
CMemosup5 1 64.1 0.30 0.30 0.5851 0.5870
time1*CMemosup5 1 78.6 2.83 2.83 0.0925 0.0965

Contrasts
Label Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
Multivariate Test of Level-1 Within-Class Emotional Support Effects 2 138 1.30 0.65 0.5214 0.5230
Multivariate Test of Level-3 Contextual Class Emotional Support Effects 2 61.2 3.91 1.96 0.1414 0.1502


PsuedoR2 (% Reduction) for CovGirlC vs. CovEmosup13C

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
CovGirlC UN(1,1) ClassID 0.02260 0.01565 1.44 0.0744 .
CovGirlC UN(2,2) ClassID 0.01259 0.006540 1.93 0.0271 .
CovGirlC UN(1,1) ClassID*StudentID 0.3907 0.04345 8.99 <.0001 .
CovGirlC UN(2,2) ClassID*StudentID 0.04893 0.01633 3.00 0.0014 .
CovGirlC Residual   0.3261 0.01942 16.79 <.0001 .
CovEmosup13C UN(1,1) ClassID 0.02660 0.01738 1.53 0.0630 -0.17679
CovEmosup13C UN(2,2) ClassID 0.004074 0.004628 0.88 0.1894 0.67655
CovEmosup13C UN(1,1) ClassID*StudentID 0.3889 0.04344 8.95 <.0001 0.00453
CovEmosup13C UN(2,2) ClassID*StudentID 0.04739 0.01632 2.90 0.0018 0.03152
CovEmosup13C Residual   0.3279 0.01956 16.76 <.0001 -0.00547


Ch 11a: Remove Class Size, Add Random Level-1 Emotional Support across Classes
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 10
Columns in Z Per Subject 53
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4650.06076811  
1 2 4141.08763083 24.04022464
2 1 4135.30138496 41.38493302
3 1 4132.55784589 53.02419197
4 1 4130.83970730 50.62999776
5 1 4128.19365604 347.60193893
6 1 4128.19100613 347.30413467
7 1 4128.18978531 347.30028430
8 1 4128.18963431 347.27432411
9 1 4128.18959655 346.37289320
10 1 4128.18959218 345.85855060
11 1 4128.18959071 354.26032390
12 2 4128.18958964 287.57882387
13 1 4128.18958816 3651.1076042
14 1 4128.05097080 82.00867561
15 1 4127.38536424 83.46869451
16 1 4127.23335877 83.28364093
17 1 4127.15852663 83.15739839
18 1 4127.12138503 83.08592127
19 1 4127.10288064 83.04809953
20 1 4127.09364478 83.02866801
21 1 4127.09133793 83.02369649
22 1 4127.09076134 83.02242352
23 1 4127.09047307 83.02183424
24 1 4127.09032893 83.02049179
25 1 4127.09025686 83.00905175
26 1 4127.09024786 83.05494639
27 1 4127.09024317 82.76605907
28 1 4127.09024140 82.38516649
29 1 4127.09024082 80.14260916
30 1 4127.09023897 67.66412514
31 3 4127.09023824 114.52102066
32 1 4127.09023430 20257.118207
33 26 4127.09023268 8542.6644380
34 1 4127.07108117 53.04768973
35 1 4126.85151318 53.83301870
36 1 4126.82292574 53.90914223
37 1 4126.81573576 53.92758896
38 1 4126.81528572 53.92911892
39 1 4126.81522941 53.93413801
40 1 4126.81522230 53.80381480
41 1 4126.81521891 53.92125272
42 1 4126.81521743 54.30030669
43 1 4126.81521528 54.73653442
44 1 4126.81520572 2137.4864055
45 10 4126.80923431 45.38997079
46 1 4126.74197806 45.65280186
47 1 4126.70772183 45.78598630
48 1 4126.69042611 45.85303383
49 1 4126.68825373 45.86135967
50 1 4126.68798201 45.86243234

WARNING: Did not converge.

Covariance Parameter Values At Last Iteration
Cov Parm Subject Estimate
UN(1,1) ClassID 0.1035
UN(2,1) ClassID 0.02874
UN(2,2) ClassID 0.04063
UN(3,1) ClassID -0.01756
UN(3,2) ClassID -0.02985
UN(3,3) ClassID 0.01113
UN(1,1) ClassID*StudentID 0.3577
UN(2,1) ClassID*StudentID 0.002871
UN(2,2) ClassID*StudentID 0.04430
Residual   0.3282


Ch 11a: New Baseline for Pseudo-R2 without Class Size Effects
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 6
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4700.00859790  
1 2 4108.71595365 0.00347774
2 2 4107.01321666 0.00037681
3 1 4106.82050311 0.00000973
4 1 4106.81585386 0.00000001
5 1 4106.81584913 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02679 0.01648 1.63 0.0520
UN(2,1) ClassID 0.02723 0.008318 3.27 0.0011
UN(2,2) ClassID 0.01583 0.007302 2.17 0.0151
UN(1,1) ClassID*StudentID 0.3906 0.04343 8.99 <.0001
UN(2,1) ClassID*StudentID -0.00524 0.02041 -0.26 0.7973
UN(2,2) ClassID*StudentID 0.04906 0.01633 3.00 0.0013
Residual   0.3260 0.01941 16.79 <.0001

Fit Statistics
-2 Res Log Likelihood 4106.8
AIC (smaller is better) 4120.8
AICC (smaller is better) 4120.9
BIC (smaller is better) 4131.3

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 593.19 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4106.8 7 4120.8 4120.9 4124.3 4131.3 4138.3

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2178 0.05598 76.8 75.34 <.0001 0.05 4.1063 4.3293
time1 -0.1273 0.03512 61.4 -3.62 0.0006 0.05 -0.1975 -0.05708
girl 0.2535 0.06880 557 3.68 0.0003 0.05 0.1183 0.3886
time1*girl -0.00639 0.03936 551 -0.16 0.8711 0.05 -0.08371 0.07093
CMgirl50 0.5512 0.4499 35.4 1.23 0.2287 0.05 -0.3619 1.4642
time1*CMgirl50 -0.8724 0.2964 33.8 -2.94 0.0058 0.05 -1.4750 -0.2699

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 61.4 13.14 13.14 0.0003 0.0006
girl 1 557 13.57 13.57 0.0002 0.0003
time1*girl 1 551 0.03 0.03 0.8711 0.8711
CMgirl50 1 35.4 1.50 1.50 0.2206 0.2287
time1*CMgirl50 1 33.8 8.66 8.66 0.0032 0.0058


Ch 11a: Add Just Level-3 Emotional Support Effects
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 8
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4642.68366010  
1 3 4102.63352458 0.00815182
2 2 4100.71068760 0.00050225
3 2 4100.47767314 0.00000407
4 1 4100.47575537 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02905 0.01775 1.64 0.0508
UN(2,1) ClassID 0.02110 0.006653 3.17 0.0015
UN(2,2) ClassID 0.004662 0.004694 0.99 0.1603
UN(1,1) ClassID*StudentID 0.3896 0.04338 8.98 <.0001
UN(2,1) ClassID*StudentID -0.00396 0.02035 -0.19 0.8459
UN(2,2) ClassID*StudentID 0.04800 0.01629 2.95 0.0016
Residual   0.3267 0.01947 16.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4100.5
AIC (smaller is better) 4114.5
AICC (smaller is better) 4114.5
BIC (smaller is better) 4125.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 542.21 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4100.5 7 4114.5 4114.5 4118.0 4125.0 4132.0

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2210 0.05658 69.2 74.60 <.0001 0.05 4.1082 4.3339
time1 -0.1298 0.02982 75.7 -4.35 <.0001 0.05 -0.1892 -0.07037
girl 0.2539 0.06878 558 3.69 0.0002 0.05 0.1188 0.3890
time1*girl -0.00768 0.03929 553 -0.20 0.8452 0.05 -0.08486 0.06950
CMgirl50 0.5293 0.4632 33 1.14 0.2614 0.05 -0.4131 1.4716
time1*CMgirl50 -1.0601 0.2372 31.8 -4.47 <.0001 0.05 -1.5434 -0.5768
CMemosup5 0.07323 0.07644 29.4 0.96 0.3459 0.05 -0.08301 0.2295
time1*CMemosup5 0.1692 0.03887 27.9 4.35 0.0002 0.05 0.08960 0.2489

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 75.7 18.94 18.94 <.0001 <.0001
girl 1 558 13.62 13.62 0.0002 0.0002
time1*girl 1 553 0.04 0.04 0.8451 0.8452
CMgirl50 1 33 1.31 1.31 0.2532 0.2614
time1*CMgirl50 1 31.8 19.97 19.97 <.0001 <.0001
CMemosup5 1 29.4 0.92 0.92 0.3381 0.3459
time1*CMemosup5 1 27.9 18.95 18.95 <.0001 0.0002

Estimates
Label Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Between-Class Emotional Support Effect at Wave 1 0.07323 0.07644 29.4 0.96 0.3459 0.05 -0.08301 0.2295
Between-Class Emotional Support Effect at Wave 2 0.2425 0.09618 28 2.52 0.0177 0.05 0.04545 0.4395
Between-Class Emotional Support Effect at Wave 3 0.4117 0.1252 27.6 3.29 0.0028 0.05 0.1550 0.6684

Contrasts
Label Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
Multivariate Test of Level-3 Between-Class Emotional Support Effects 2 31.1 19.16 9.58 <.0001 0.0006


PsuedoR2 (% Reduction) for FitGirlNewC vs. CovEmosup3C

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
FitGirlNewC UN(1,1) ClassID 0.02679 0.01648 1.63 0.0520 .
FitGirlNewC UN(2,2) ClassID 0.01583 0.007302 2.17 0.0151 .
FitGirlNewC UN(1,1) ClassID*StudentID 0.3906 0.04343 8.99 <.0001 .
FitGirlNewC UN(2,2) ClassID*StudentID 0.04906 0.01633 3.00 0.0013 .
FitGirlNewC Residual   0.3260 0.01941 16.79 <.0001 .
CovEmosup3C UN(1,1) ClassID 0.02905 0.01775 1.64 0.0508 -0.08416
CovEmosup3C UN(2,2) ClassID 0.004662 0.004694 0.99 0.1603 0.70541
CovEmosup3C UN(1,1) ClassID*StudentID 0.3896 0.04338 8.98 <.0001 0.00247
CovEmosup3C UN(2,2) ClassID*StudentID 0.04800 0.01629 2.95 0.0016 0.02175
CovEmosup3C Residual   0.3267 0.01947 16.78 <.0001 -0.00235


Eq 11a.3: Saturated Means, Unstructured Level-3 and Level-2 Variances
for Student Victimization

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable victim
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method None
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 12
Columns in X 4
Columns in Z Per Subject 3
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4890.85920241  
1 2 4155.11463166 0.00227367
2 1 4154.11673467 0.00014077
3 1 4154.04382302 0.00000138
4 1 4154.04314069 0.00000000

Convergence criteria met.

Estimated R Matrix for ClassID*StudentID
130231 5023101
Row Col1 Col2 Col3
1 0.8834 0.6243 0.5605
2 0.6243 1.0204 0.6505
3 0.5605 0.6505 0.9397

Estimated R Correlation Matrix for
ClassID*StudentID 130231 5023101
Row Col1 Col2 Col3
1 1.0000 0.6576 0.6151
2 0.6576 1.0000 0.6643
3 0.6151 0.6643 1.0000

Estimated G Matrix
Row Effect ClassID: Class
ID Variable		 wave:
Wave of
Study
(1-3)		 Col1 Col2 Col3
1 wave 130231 1 0.02297 0.02452 0.02641
2 wave 130231 2 0.02452 0.05279 0.03961
3 wave 130231 3 0.02641 0.03961 0.03345

Estimated G Correlation Matrix
Row Effect ClassID: Class
ID Variable		 wave:
Wave of
Study
(1-3)		 Col1 Col2 Col3
1 wave 130231 1 1.0000 0.7043 0.9528
2 wave 130231 2 0.7043 1.0000 0.9425
3 wave 130231 3 0.9528 0.9425 1.0000

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02297 0.01921 1.20 0.1160
UN(2,1) ClassID 0.02452 0.01965 1.25 0.2121
UN(2,2) ClassID 0.05279 0.02786 1.89 0.0291
UN(3,1) ClassID 0.02641 0.01828 1.44 0.1486
UN(3,2) ClassID 0.03961 0.02237 1.77 0.0767
UN(3,3) ClassID 0.03345 0.02256 1.48 0.0691
UN(1,1) ClassID*StudentID 0.8834 0.05362 16.48 <.0001
UN(2,1) ClassID*StudentID 0.6243 0.04841 12.90 <.0001
UN(2,2) ClassID*StudentID 1.0204 0.06112 16.69 <.0001
UN(3,1) ClassID*StudentID 0.5605 0.04605 12.17 <.0001
UN(3,2) ClassID*StudentID 0.6505 0.05010 12.98 <.0001
UN(3,3) ClassID*StudentID 0.9397 0.05710 16.46 <.0001

Fit Statistics
-2 Res Log Likelihood 4154.0
AIC (smaller is better) 4178.0
AICC (smaller is better) 4178.2
BIC (smaller is better) 4196.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
11 736.82 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4154.0 12 4178.0 4178.2 4184.1 4196.0 4208.0

Solution for Fixed Effects
Effect wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept   3.1854 0.05131 29.6 62.08 <.0001 0.05 3.0805 3.2902
wave 1 -0.1056 0.03676 33.9 -2.87 0.0070 0.05 -0.1803 -0.03090
wave 2 -0.06068 0.03700 27.6 -1.64 0.1124 0.05 -0.1365 0.01517
wave 3 0 . . . . . . .

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
wave 2 30.5 9.04 4.52 0.0109 0.0191

Least Squares Means
Effect wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
wave 1 3.0798 0.04717 29.2 65.29 <.0001 0.05 2.9833 3.1762
wave 2 3.1247 0.05794 31.5 53.93 <.0001 0.05 3.0066 3.2428
wave 3 3.1854 0.05131 29.6 62.08 <.0001 0.05 3.0805 3.2902

Differences of Least Squares Means
Effect wave:
Wave of
Study
(1-3)		 wave:
Wave of
Study
(1-3)		 Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
wave 1 2 -0.04493 0.04443 32.7 -1.01 0.3193 0.05 -0.1354 0.04550
wave 1 3 -0.1056 0.03676 33.9 -2.87 0.0070 0.05 -0.1803 -0.03090
wave 2 3 -0.06068 0.03700 27.6 -1.64 0.1124 0.05 -0.1365 0.01517


Ch 11a: Piecewise Means, Level-3 and Level-2 Random Intercepts
Three-Level Model Predicting Student Victimization

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable victim
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 3
Columns in X 3
Columns in Z Per Subject 26
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4890.85920241  
1 2 4169.54164794 0.00000180
2 1 4169.54074302 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID 0.03296 0.01899 1.74 0.0413
UN(1,1) ClassID*StudentID 0.6110 0.04386 13.93 <.0001
Residual   0.3419 0.01437 23.79 <.0001

Fit Statistics
-2 Res Log Likelihood 4169.5
AIC (smaller is better) 4175.5
AICC (smaller is better) 4175.6
BIC (smaller is better) 4180.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 721.32 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4169.5 3 4175.5 4175.6 4177.1 4180.0 4183.0

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 3.0820 0.05147 42.3 59.88 <.0001 0.05 2.9781 3.1858
time1 0.04227 0.03462 1138 1.22 0.2223 0.05 -0.02566 0.1102
w3 0.01856 0.05976 1138 0.31 0.7562 0.05 -0.09870 0.1358

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 1138 1.49 1.49 0.2221 0.2223
w3 1 1138 0.10 0.10 0.7561 0.7562


Likelihood Ratio Test for FitPieceRI2RI3V vs. FitSatUNV

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitPieceRI2RI3V 4169.5 3 4175.5 4180.0 . . .
FitSatUNV 4154.0 12 4178.0 4196.0 15.4976 9 0.078144


Ch 11a: Piecewise Means, Add Level-2 Random Time Slope
Three-Level Model Predicting Student Victimization

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable victim
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 5
Columns in X 3
Columns in Z Per Subject 51
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4890.85920241  
1 3 4167.90029045 0.00000117
2 1 4167.89970797 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03220 0.01881 1.71 0.0434
UN(1,1) ClassID*StudentID 0.5925 0.05477 10.82 <.0001
UN(2,1) ClassID*StudentID 0.006145 0.02069 0.30 0.7664
UN(2,2) ClassID*StudentID 0.01019 0.01443 0.71 0.2399
Residual   0.3317 0.01967 16.87 <.0001

Fit Statistics
-2 Res Log Likelihood 4167.9
AIC (smaller is better) 4177.9
AICC (smaller is better) 4177.9
BIC (smaller is better) 4185.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
4 722.96 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4167.9 5 4177.9 4177.9 4180.4 4185.4 4190.4

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 3.0820 0.05077 40.3 60.71 <.0001 0.05 2.9794 3.1846
time1 0.04245 0.03436 946 1.24 0.2171 0.05 -0.02499 0.1099
w3 0.01848 0.05889 586 0.31 0.7537 0.05 -0.09718 0.1341

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 946 1.53 1.53 0.2168 0.2171
w3 1 586 0.10 0.10 0.7536 0.7537


Likelihood Ratio Test for FitPieceRI2RI3V vs. FitPieceRL2RI3V

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitPieceRI2RI3V 4169.5 3 4175.5 4180.0 . . .
FitPieceRL2RI3V 4167.9 5 4177.9 4185.4 1.64104 2 0.44020


Ch 11a: Piecewise Means, Add Level-3 Random Time Slope
Three-Level Model Predicting Student Victimization

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable victim
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 3
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4890.85920241  
1 4 4167.22054709 .
2 1 4166.95007519 0.00002040
3 1 4166.93962997 0.00000005
4 1 4166.93960285 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02484 0.01930 1.29 0.0990
UN(2,1) ClassID 0.003351 0.005332 0.63 0.5297
UN(2,2) ClassID 0.001214 0.002781 0.44 0.3312
UN(1,1) ClassID*StudentID 0.5961 0.05545 10.75 <.0001
UN(2,1) ClassID*StudentID 0.004837 0.02102 0.23 0.8180
UN(2,2) ClassID*StudentID 0.009060 0.01461 0.62 0.2676
Residual   0.3317 0.01966 16.87 <.0001

Fit Statistics
-2 Res Log Likelihood 4166.9
AIC (smaller is better) 4180.9
AICC (smaller is better) 4181.0
BIC (smaller is better) 4191.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 723.92 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4166.9 7 4180.9 4181.0 4184.5 4191.4 4198.4

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 3.0810 0.04855 32.5 63.46 <.0001 0.05 2.9821 3.1798
time1 0.04228 0.03488 302 1.21 0.2264 0.05 -0.02636 0.1109
w3 0.01914 0.05888 586 0.32 0.7453 0.05 -0.09651 0.1348

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 302 1.47 1.47 0.2254 0.2264
w3 1 586 0.11 0.11 0.7452 0.7453


Likelihood Ratio Test for FitPieceRL2RI3V vs. FitPieceRL2RL3V

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitPieceRL2RI3V 4167.9 5 4177.9 4185.4 . . .
FitPieceRL2RL3V 4166.9 7 4180.9 4191.4 0.96011 2 0.61875


Likelihood Ratio Test for FitPieceRL2RL3V vs. FitSatUNV

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitPieceRL2RL3V 4166.9 7 4180.9 4191.4 . . .
FitSatUNV 4154.0 12 4178.0 4196.0 12.8965 5 0.024369


Ch 11a: Fixed Linear Time, Level-3 and Level-2 Random Intercepts
Three-Level Model Predicting Student Victimization

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable victim
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 3
Columns in X 2
Columns in Z Per Subject 26
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4888.11029571  
1 2 4165.84092185 0.00000182
2 1 4165.84001353 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr > Z
UN(1,1) ClassID 0.03298 0.01899 1.74 0.0412
UN(1,1) ClassID*StudentID 0.6111 0.04386 13.93 <.0001
Residual   0.3417 0.01436 23.80 <.0001

Fit Statistics
-2 Res Log Likelihood 4165.8
AIC (smaller is better) 4171.8
AICC (smaller is better) 4171.9
BIC (smaller is better) 4176.3

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
2 722.27 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4165.8 3 4171.8 4171.9 4173.4 4176.3 4179.3

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 3.0788 0.05047 39.1 61.00 <.0001 0.05 2.9768 3.1809
time1 0.05157 0.01740 1144 2.96 0.0031 0.05 0.01743 0.08571

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 1144 8.78 8.78 0.0030 0.0031


Ch 11a: Add All 3 Main Effects of Student Vicitimization Predicting Student Closeness
Using Variable-Centered Level-1 and Level-2 Victim Predictors

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 11
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4631.64339684  
1 3 4101.17076438 0.00846271
2 2 4099.22166881 0.00051611
3 2 4098.98058489 0.00000397
4 1 4098.97870474 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03161 0.01853 1.71 0.0440
UN(2,1) ClassID 0.02142 0.006742 3.18 0.0015
UN(2,2) ClassID 0.004514 0.004678 0.97 0.1673
UN(1,1) ClassID*StudentID 0.3811 0.04299 8.86 <.0001
UN(2,1) ClassID*StudentID -0.00596 0.02035 -0.29 0.7695
UN(2,2) ClassID*StudentID 0.04847 0.01633 2.97 0.0015
Residual   0.3268 0.01948 16.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4099.0
AIC (smaller is better) 4113.0
AICC (smaller is better) 4113.0
BIC (smaller is better) 4123.5

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 532.66 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4099.0 7 4113.0 4113.0 4116.5 4123.5 4130.5

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2229 0.06133 59.5 68.86 <.0001 0.05 4.1003 4.3456
time1 -0.1304 0.02985 76.6 -4.37 <.0001 0.05 -0.1898 -0.07092
girl 0.2581 0.06837 556 3.78 0.0002 0.05 0.1238 0.3924
time1*girl -0.00722 0.03935 553 -0.18 0.8546 0.05 -0.08450 0.07007
CMgirl50 0.5160 0.4852 32.4 1.06 0.2954 0.05 -0.4718 1.5038
time1*CMgirl50 -1.0621 0.2365 31.7 -4.49 <.0001 0.05 -1.5439 -0.5802
CMemosup5 0.06855 0.08031 29.7 0.85 0.4002 0.05 -0.09553 0.2326
time1*CMemosup5 0.1700 0.03876 27.8 4.39 0.0002 0.05 0.09055 0.2494
WSvictim 0.01087 0.03072 1113 0.35 0.7234 0.05 -0.04940 0.07115
WCvictim -0.1292 0.03568 559 -3.62 0.0003 0.05 -0.1993 -0.05909
CMvictim3 -0.02480 0.1747 34 -0.14 0.8880 0.05 -0.3799 0.3303

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 76.6 19.07 19.07 <.0001 <.0001
girl 1 556 14.25 14.25 0.0002 0.0002
time1*girl 1 553 0.03 0.03 0.8545 0.8546
CMgirl50 1 32.4 1.13 1.13 0.2875 0.2954
time1*CMgirl50 1 31.7 20.17 20.17 <.0001 <.0001
CMemosup5 1 29.7 0.73 0.73 0.3934 0.4002
time1*CMemosup5 1 27.8 19.23 19.23 <.0001 0.0002
WSvictim 1 1113 0.13 0.13 0.7234 0.7234
WCvictim 1 559 13.11 13.11 0.0003 0.0003
CMvictim3 1 34 0.02 0.02 0.8871 0.8880

Estimates
Label Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Level-1 Within-Student Victim Effect 0.01087 0.03072 1113 0.35 0.7234 0.05 -0.04940 0.07115
Level-2 Within-Class Victim Effect -0.1292 0.03568 559 -3.62 0.0003 0.05 -0.1993 -0.05909
Level-3 Between-Class Victim Effect -0.02480 0.1747 34 -0.14 0.8880 0.05 -0.3799 0.3303
Level-2 Within-Class Contextual Victim Effect -0.1401 0.04699 1308 -2.98 0.0029 0.05 -0.2322 -0.04787
Level-3 Between-Class Contextual Victim Effect 0.1044 0.1783 36.8 0.59 0.5619 0.05 -0.2570 0.4658


PsuedoR2 (% Reduction) for CovEmosup3C vs. CovVicVBC

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
CovEmosup3C UN(1,1) ClassID 0.02905 0.01775 1.64 0.0508 .
CovEmosup3C UN(2,2) ClassID 0.004662 0.004694 0.99 0.1603 .
CovEmosup3C UN(1,1) ClassID*StudentID 0.3896 0.04338 8.98 <.0001 .
CovEmosup3C UN(2,2) ClassID*StudentID 0.04800 0.01629 2.95 0.0016 .
CovEmosup3C Residual   0.3267 0.01947 16.78 <.0001 .
CovVicVBC UN(1,1) ClassID 0.03161 0.01853 1.71 0.0440 -0.088310
CovVicVBC UN(2,2) ClassID 0.004514 0.004678 0.97 0.1673 0.031731
CovVicVBC UN(1,1) ClassID*StudentID 0.3811 0.04299 8.86 <.0001 0.021775
CovVicVBC UN(2,2) ClassID*StudentID 0.04847 0.01633 2.97 0.0015 -0.009905
CovVicVBC Residual   0.3268 0.01948 16.78 <.0001 -0.000132


Ch 11a: Add All 3 Main Effects of Student Vicitimization Predicting Student Closeness
Using Constant-Centered Level-1 and Level-2 Victim Predictors

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 11
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4631.64339684  
1 3 4101.17076438 0.00846271
2 2 4099.22166881 0.00051611
3 2 4098.98058489 0.00000397
4 1 4098.97870474 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03161 0.01853 1.71 0.0440
UN(2,1) ClassID 0.02142 0.006742 3.18 0.0015
UN(2,2) ClassID 0.004514 0.004678 0.97 0.1673
UN(1,1) ClassID*StudentID 0.3811 0.04299 8.86 <.0001
UN(2,1) ClassID*StudentID -0.00596 0.02035 -0.29 0.7695
UN(2,2) ClassID*StudentID 0.04847 0.01633 2.97 0.0015
Residual   0.3268 0.01948 16.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4099.0
AIC (smaller is better) 4113.0
AICC (smaller is better) 4113.0
BIC (smaller is better) 4123.5

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 532.66 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4099.0 7 4113.0 4113.0 4116.5 4123.5 4130.5

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2229 0.06133 59.5 68.86 <.0001 0.05 4.1003 4.3456
time1 -0.1304 0.02985 76.6 -4.37 <.0001 0.05 -0.1898 -0.07092
girl 0.2581 0.06837 556 3.78 0.0002 0.05 0.1238 0.3924
time1*girl -0.00722 0.03935 553 -0.18 0.8546 0.05 -0.08450 0.07007
CMgirl50 0.5160 0.4852 32.4 1.06 0.2954 0.05 -0.4718 1.5038
time1*CMgirl50 -1.0621 0.2365 31.7 -4.49 <.0001 0.05 -1.5439 -0.5802
CMemosup5 0.06855 0.08031 29.7 0.85 0.4002 0.05 -0.09553 0.2326
time1*CMemosup5 0.1700 0.03876 27.8 4.39 0.0002 0.05 0.09055 0.2494
victim3 0.01087 0.03072 1113 0.35 0.7234 0.05 -0.04940 0.07115
SMvictim3 -0.1401 0.04699 1308 -2.98 0.0029 0.05 -0.2322 -0.04787
CMvictim3 0.1044 0.1783 36.8 0.59 0.5619 0.05 -0.2570 0.4658

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 76.6 19.07 19.07 <.0001 <.0001
girl 1 556 14.25 14.25 0.0002 0.0002
time1*girl 1 553 0.03 0.03 0.8545 0.8546
CMgirl50 1 32.4 1.13 1.13 0.2875 0.2954
time1*CMgirl50 1 31.7 20.17 20.17 <.0001 <.0001
CMemosup5 1 29.7 0.73 0.73 0.3934 0.4002
time1*CMemosup5 1 27.8 19.23 19.23 <.0001 0.0002
victim3 1 1113 0.13 0.13 0.7234 0.7234
SMvictim3 1 1308 8.88 8.88 0.0029 0.0029
CMvictim3 1 36.8 0.34 0.34 0.5583 0.5619

Estimates
Label Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Level-1 Within-Student Victim Effect 0.01087 0.03072 1113 0.35 0.7234 0.05 -0.04940 0.07115
Level-2 Within-Class Victim Effect -0.1292 0.03568 559 -3.62 0.0003 0.05 -0.1993 -0.05909
Level-3 Between-Class Victim Effect -0.02480 0.1747 34 -0.14 0.8880 0.05 -0.3799 0.3303
Level-2 Within-Class Contextual Victim Effect -0.1401 0.04699 1308 -2.98 0.0029 0.05 -0.2322 -0.04787
Level-3 Between-Class Contextual Victim Effect 0.1044 0.1783 36.8 0.59 0.5619 0.05 -0.2570 0.4658


PsuedoR2 (% Reduction) for CovEmosup3C vs. CovVicCBC

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
CovEmosup3C UN(1,1) ClassID 0.02905 0.01775 1.64 0.0508 .
CovEmosup3C UN(2,2) ClassID 0.004662 0.004694 0.99 0.1603 .
CovEmosup3C UN(1,1) ClassID*StudentID 0.3896 0.04338 8.98 <.0001 .
CovEmosup3C UN(2,2) ClassID*StudentID 0.04800 0.01629 2.95 0.0016 .
CovEmosup3C Residual   0.3267 0.01947 16.78 <.0001 .
CovVicCBC UN(1,1) ClassID 0.03161 0.01853 1.71 0.0440 -0.088310
CovVicCBC UN(2,2) ClassID 0.004514 0.004678 0.97 0.1673 0.031731
CovVicCBC UN(1,1) ClassID*StudentID 0.3811 0.04299 8.86 <.0001 0.021775
CovVicCBC UN(2,2) ClassID*StudentID 0.04847 0.01633 2.97 0.0015 -0.009905
CovVicCBC Residual   0.3268 0.01948 16.78 <.0001 -0.000132


Ch 11a: Constant-Centered Student Vicitimization Predicting Student Closeness
For Table 11.3: Level-2 and Level-3 Effects, Omitting Level-1 Effect

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 10
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4627.37814710  
1 3 4096.13797924 0.00831321
2 2 4094.20788925 0.00049797
3 2 4093.97663081 0.00000366
4 1 4093.97491349 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03160 0.01853 1.71 0.0440
UN(2,1) ClassID 0.02136 0.006748 3.17 0.0015
UN(2,2) ClassID 0.004601 0.004685 0.98 0.1630
UN(1,1) ClassID*StudentID 0.3810 0.04297 8.87 <.0001
UN(2,1) ClassID*StudentID -0.00578 0.02033 -0.28 0.7763
UN(2,2) ClassID*StudentID 0.04831 0.01630 2.96 0.0015
Residual   0.3266 0.01946 16.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4094.0
AIC (smaller is better) 4108.0
AICC (smaller is better) 4108.0
BIC (smaller is better) 4118.5

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 533.40 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4094.0 7 4108.0 4108.0 4111.5 4118.5 4125.5

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2222 0.06129 59.3 68.89 <.0001 0.05 4.0996 4.3449
time1 -0.1296 0.02980 75.8 -4.35 <.0001 0.05 -0.1890 -0.07024
girl 0.2585 0.06834 556 3.78 0.0002 0.05 0.1242 0.3927
time1*girl -0.00755 0.03931 553 -0.19 0.8478 0.05 -0.08477 0.06967
CMgirl50 0.5163 0.4851 32.4 1.06 0.2951 0.05 -0.4714 1.5040
time1*CMgirl50 -1.0629 0.2369 31.8 -4.49 <.0001 0.05 -1.5456 -0.5803
CMemosup5 0.06891 0.08029 29.7 0.86 0.3977 0.05 -0.09515 0.2330
time1*CMemosup5 0.1696 0.03882 27.9 4.37 0.0002 0.05 0.09003 0.2491
SMvictim3 -0.1292 0.03568 559 -3.62 0.0003 0.05 -0.1993 -0.05914
CMvictim3 0.1038 0.1784 36.8 0.58 0.5643 0.05 -0.2577 0.4653

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 75.8 18.91 18.91 <.0001 <.0001
girl 1 556 14.30 14.30 0.0002 0.0002
time1*girl 1 553 0.04 0.04 0.8477 0.8478
CMgirl50 1 32.4 1.13 1.13 0.2872 0.2951
time1*CMgirl50 1 31.8 20.13 20.13 <.0001 <.0001
CMemosup5 1 29.7 0.74 0.74 0.3908 0.3977
time1*CMemosup5 1 27.9 19.08 19.08 <.0001 0.0002
SMvictim3 1 559 13.11 13.11 0.0003 0.0003
CMvictim3 1 36.8 0.34 0.34 0.5607 0.5643


Ch 11a: Constant-Centered Student Vicitimization Predicting Student Closeness
For Table 11.3: Level-1 and Level-3 Effects Only, Omitting Level-2 Effect

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 10
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4634.22342538  
1 3 4105.92376359 0.00860744
2 2 4103.84782352 0.00067348
3 2 4103.52552999 0.00000476
4 1 4103.52327383 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03129 0.01850 1.69 0.0454
UN(2,1) ClassID 0.02102 0.006786 3.10 0.0020
UN(2,2) ClassID 0.004995 0.004767 1.05 0.1474
UN(1,1) ClassID*StudentID 0.3824 0.04322 8.85 <.0001
UN(2,1) ClassID*StudentID -0.00403 0.02035 -0.20 0.8431
UN(2,2) ClassID*StudentID 0.04750 0.01634 2.91 0.0018
Residual   0.3283 0.01959 16.76 <.0001

Fit Statistics
-2 Res Log Likelihood 4103.5
AIC (smaller is better) 4117.5
AICC (smaller is better) 4117.6
BIC (smaller is better) 4128.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 530.70 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4103.5 7 4117.5 4117.6 4121.0 4128.0 4135.0

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2206 0.06135 59.4 68.79 <.0001 0.05 4.0978 4.3433
time1 -0.1264 0.03006 75.7 -4.20 <.0001 0.05 -0.1863 -0.06652
girl 0.2572 0.06850 555 3.75 0.0002 0.05 0.1226 0.3917
time1*girl -0.00911 0.03933 553 -0.23 0.8169 0.05 -0.08636 0.06814
CMgirl50 0.5160 0.4849 32.4 1.06 0.2951 0.05 -0.4712 1.5033
time1*CMgirl50 -1.0658 0.2396 32.1 -4.45 <.0001 0.05 -1.5537 -0.5779
CMemosup5 0.07106 0.08024 29.7 0.89 0.3830 0.05 -0.09289 0.2350
time1*CMemosup5 0.1678 0.03928 28.2 4.27 0.0002 0.05 0.08733 0.2482
victim3 -0.04877 0.02339 1612 -2.09 0.0372 0.05 -0.09464 -0.00289
CMvictim3 0.02064 0.1767 34.7 0.12 0.9077 0.05 -0.3381 0.3794

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 75.7 17.68 17.68 <.0001 <.0001
girl 1 555 14.10 14.10 0.0002 0.0002
time1*girl 1 553 0.05 0.05 0.8168 0.8169
CMgirl50 1 32.4 1.13 1.13 0.2872 0.2951
time1*CMgirl50 1 32.1 19.79 19.79 <.0001 <.0001
CMemosup5 1 29.7 0.78 0.78 0.3759 0.3830
time1*CMemosup5 1 28.2 18.24 18.24 <.0001 0.0002
victim3 1 1612 4.35 4.35 0.0371 0.0372
CMvictim3 1 34.7 0.01 0.01 0.9070 0.9077


Ch 11a: Constant-Centered Student Vicitimization Predicting Student Closeness
For Table 11.3: Level-1 and Level-2 Effects, Omitting Level-3 Effect

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 10
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4629.71652035  
1 3 4100.02330569 0.00926617
2 2 4097.96117264 0.00055269
3 2 4097.70507661 0.00000505
4 1 4097.70268694 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03013 0.01795 1.68 0.0467
UN(2,1) ClassID 0.02134 0.006678 3.20 0.0014
UN(2,2) ClassID 0.004532 0.004679 0.97 0.1664
UN(1,1) ClassID*StudentID 0.3811 0.04299 8.87 <.0001
UN(2,1) ClassID*StudentID -0.00590 0.02035 -0.29 0.7720
UN(2,2) ClassID*StudentID 0.04844 0.01632 2.97 0.0015
Residual   0.3268 0.01948 16.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4097.7
AIC (smaller is better) 4111.7
AICC (smaller is better) 4111.8
BIC (smaller is better) 4122.2

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 532.01 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4097.7 7 4111.7 4111.8 4115.2 4122.2 4129.2

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2357 0.05678 68.6 74.59 <.0001 0.05 4.1224 4.3490
time1 -0.1301 0.02985 76.6 -4.36 <.0001 0.05 -0.1895 -0.07064
girl 0.2581 0.06837 556 3.77 0.0002 0.05 0.1238 0.3924
time1*girl -0.00723 0.03934 553 -0.18 0.8542 0.05 -0.08451 0.07005
CMgirl50 0.4446 0.4655 33 0.96 0.3465 0.05 -0.5026 1.3918
time1*CMgirl50 -1.0618 0.2365 31.7 -4.49 <.0001 0.05 -1.5438 -0.5798
CMemosup5 0.05708 0.07690 29.5 0.74 0.4638 0.05 -0.1001 0.2143
time1*CMemosup5 0.1702 0.03877 27.8 4.39 0.0001 0.05 0.09074 0.2496
victim3 0.01073 0.03071 1112 0.35 0.7268 0.05 -0.04953 0.07099
SMvictim3 -0.1357 0.04638 1361 -2.92 0.0035 0.05 -0.2267 -0.04467

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 76.6 18.99 18.99 <.0001 <.0001
girl 1 556 14.25 14.25 0.0002 0.0002
time1*girl 1 553 0.03 0.03 0.8542 0.8542
CMgirl50 1 33 0.91 0.91 0.3396 0.3465
time1*CMgirl50 1 31.7 20.15 20.15 <.0001 <.0001
CMemosup5 1 29.5 0.55 0.55 0.4579 0.4638
time1*CMemosup5 1 27.8 19.27 19.27 <.0001 0.0001
victim3 1 1112 0.12 0.12 0.7267 0.7268
SMvictim3 1 1361 8.55 8.55 0.0034 0.0035


Ch 11a: Constant-Centered Student Vicitimization Predicting Student Closeness
For Table 11.3: Level-1 Effect Only, Omitting Level-2 and Level-3 Effects

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 9
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4631.71335693  
1 3 4104.37913211 0.00909419
2 2 4102.22358601 0.00069133
3 2 4101.89577322 0.00000549
4 1 4101.89317939 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02931 0.01775 1.65 0.0493
UN(2,1) ClassID 0.02092 0.006697 3.12 0.0018
UN(2,2) ClassID 0.005042 0.004778 1.06 0.1457
UN(1,1) ClassID*StudentID 0.3823 0.04320 8.85 <.0001
UN(2,1) ClassID*StudentID -0.00394 0.02034 -0.19 0.8465
UN(2,2) ClassID*StudentID 0.04745 0.01633 2.91 0.0018
Residual   0.3283 0.01959 16.76 <.0001

Fit Statistics
-2 Res Log Likelihood 4101.9
AIC (smaller is better) 4115.9
AICC (smaller is better) 4116.0
BIC (smaller is better) 4126.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 529.82 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4101.9 7 4115.9 4116.0 4119.4 4126.4 4133.4

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2230 0.05648 68.7 74.76 <.0001 0.05 4.1103 4.3357
time1 -0.1263 0.03008 75.6 -4.20 <.0001 0.05 -0.1862 -0.06638
girl 0.2572 0.06849 556 3.75 0.0002 0.05 0.1226 0.3917
time1*girl -0.00909 0.03932 553 -0.23 0.8173 0.05 -0.08632 0.06815
CMgirl50 0.4996 0.4631 33 1.08 0.2885 0.05 -0.4426 1.4418
time1*CMgirl50 -1.0650 0.2398 32.1 -4.44 <.0001 0.05 -1.5534 -0.5766
CMemosup5 0.06914 0.07643 29.4 0.90 0.3731 0.05 -0.08709 0.2254
time1*CMemosup5 0.1677 0.03932 28.2 4.26 0.0002 0.05 0.08716 0.2482
victim3 -0.04838 0.02320 1626 -2.08 0.0372 0.05 -0.09389 -0.00287

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 75.6 17.63 17.63 <.0001 <.0001
girl 1 556 14.10 14.10 0.0002 0.0002
time1*girl 1 553 0.05 0.05 0.8172 0.8173
CMgirl50 1 33 1.16 1.16 0.2807 0.2885
time1*CMgirl50 1 32.1 19.72 19.72 <.0001 <.0001
CMemosup5 1 29.4 0.82 0.82 0.3657 0.3731
time1*CMemosup5 1 28.2 18.18 18.18 <.0001 0.0002
victim3 1 1626 4.35 4.35 0.0371 0.0372


Ch 11a: Add All 3 Victim*Time Interactions Predicting Student Closeness
Using Variable-Centered Level-1 and Level-2 Victim Predictors

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 14
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4642.95969911  
1 3 4113.13174189 0.00667741
2 2 4111.41066577 0.00035610
3 2 4111.23932711 0.00000163
4 1 4111.23854576 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03174 0.01867 1.70 0.0445
UN(2,1) ClassID 0.02194 0.007018 3.13 0.0018
UN(2,2) ClassID 0.005182 0.004960 1.04 0.1481
UN(1,1) ClassID*StudentID 0.3816 0.04303 8.87 <.0001
UN(2,1) ClassID*StudentID -0.00631 0.02039 -0.31 0.7571
UN(2,2) ClassID*StudentID 0.04893 0.01636 2.99 0.0014
Residual   0.3268 0.01948 16.77 <.0001

Fit Statistics
-2 Res Log Likelihood 4111.2
AIC (smaller is better) 4125.2
AICC (smaller is better) 4125.3
BIC (smaller is better) 4135.7

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 531.72 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4111.2 7 4125.2 4125.3 4128.8 4135.7 4142.7

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2239 0.06193 55.5 68.20 <.0001 0.05 4.0999 4.3480
time1 -0.1293 0.03249 59.3 -3.98 0.0002 0.05 -0.1943 -0.06428
girl 0.2578 0.06843 556 3.77 0.0002 0.05 0.1234 0.3922
time1*girl -0.00671 0.03941 552 -0.17 0.8648 0.05 -0.08413 0.07070
CMgirl50 0.5097 0.4875 31.6 1.05 0.3037 0.05 -0.4839 1.5033
time1*CMgirl50 -1.0637 0.2493 30.1 -4.27 0.0002 0.05 -1.5726 -0.5547
CMemosup5 0.06853 0.08061 29 0.85 0.4022 0.05 -0.09633 0.2334
time1*CMemosup5 0.1681 0.04106 27.4 4.09 0.0003 0.05 0.08390 0.2523
WSvictim 0.01000 0.05585 1255 0.18 0.8579 0.05 -0.09956 0.1196
WCvictim -0.1151 0.04045 559 -2.85 0.0046 0.05 -0.1945 -0.03563
CMvictim3 -0.03034 0.1838 29.8 -0.17 0.8700 0.05 -0.4059 0.3452
time1*WSvictim 0.001416 0.04816 1504 0.03 0.9766 0.05 -0.09305 0.09588
time1*WCvictim -0.01754 0.02355 564 -0.74 0.4567 0.05 -0.06379 0.02871
time1*CMvictim3 -0.00908 0.09288 27.5 -0.10 0.9228 0.05 -0.1995 0.1813

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 59.3 15.83 15.83 <.0001 0.0002
girl 1 556 14.19 14.19 0.0002 0.0002
time1*girl 1 552 0.03 0.03 0.8647 0.8648
CMgirl50 1 31.6 1.09 1.09 0.2958 0.3037
time1*CMgirl50 1 30.1 18.21 18.21 <.0001 0.0002
CMemosup5 1 29 0.72 0.72 0.3952 0.4022
time1*CMemosup5 1 27.4 16.76 16.76 <.0001 0.0003
WSvictim 1 1255 0.03 0.03 0.8578 0.8579
WCvictim 1 559 8.09 8.09 0.0044 0.0046
CMvictim3 1 29.8 0.03 0.03 0.8689 0.8700
time1*WSvictim 1 1504 0.00 0.00 0.9765 0.9766
time1*WCvictim 1 564 0.55 0.55 0.4564 0.4567
time1*CMvictim3 1 27.5 0.01 0.01 0.9221 0.9228


Ch 11a: Add All 3 Victim*Time Interactions Predicting Student Closeness
Using Constant-Centered Level-1 and Level-2 Victim Predictors

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 7
Columns in X 14
Columns in Z Per Subject 52
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4642.95969911  
1 3 4113.13174189 0.00667741
2 2 4111.41066577 0.00035610
3 2 4111.23932711 0.00000163
4 1 4111.23854576 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03174 0.01867 1.70 0.0445
UN(2,1) ClassID 0.02194 0.007018 3.13 0.0018
UN(2,2) ClassID 0.005182 0.004960 1.04 0.1481
UN(1,1) ClassID*StudentID 0.3816 0.04303 8.87 <.0001
UN(2,1) ClassID*StudentID -0.00631 0.02039 -0.31 0.7571
UN(2,2) ClassID*StudentID 0.04893 0.01636 2.99 0.0014
Residual   0.3268 0.01948 16.77 <.0001

Fit Statistics
-2 Res Log Likelihood 4111.2
AIC (smaller is better) 4125.2
AICC (smaller is better) 4125.3
BIC (smaller is better) 4135.7

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
6 531.72 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4111.2 7 4125.2 4125.3 4128.8 4135.7 4142.7

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2239 0.06193 55.5 68.20 <.0001 0.05 4.0999 4.3480
time1 -0.1293 0.03249 59.3 -3.98 0.0002 0.05 -0.1943 -0.06428
girl 0.2578 0.06843 556 3.77 0.0002 0.05 0.1234 0.3922
time1*girl -0.00671 0.03941 552 -0.17 0.8648 0.05 -0.08413 0.07070
CMgirl50 0.5097 0.4875 31.6 1.05 0.3037 0.05 -0.4839 1.5033
time1*CMgirl50 -1.0637 0.2493 30.1 -4.27 0.0002 0.05 -1.5726 -0.5547
CMemosup5 0.06853 0.08061 29 0.85 0.4022 0.05 -0.09633 0.2334
time1*CMemosup5 0.1681 0.04106 27.4 4.09 0.0003 0.05 0.08390 0.2523
victim3 0.01000 0.05585 1255 0.18 0.8579 0.05 -0.09956 0.1196
SMvictim3 -0.1251 0.06731 1184 -1.86 0.0634 0.05 -0.2572 0.006977
CMvictim3 0.08475 0.1882 32.7 0.45 0.6554 0.05 -0.2983 0.4678
time1*victim3 0.001416 0.04816 1504 0.03 0.9766 0.05 -0.09305 0.09588
time1*SMvictim3 -0.01895 0.05245 1458 -0.36 0.7179 0.05 -0.1218 0.08393
time1*CMvictim3 0.008458 0.09578 31.1 0.09 0.9302 0.05 -0.1869 0.2038

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 59.3 15.83 15.83 <.0001 0.0002
girl 1 556 14.19 14.19 0.0002 0.0002
time1*girl 1 552 0.03 0.03 0.8647 0.8648
CMgirl50 1 31.6 1.09 1.09 0.2958 0.3037
time1*CMgirl50 1 30.1 18.21 18.21 <.0001 0.0002
CMemosup5 1 29 0.72 0.72 0.3952 0.4022
time1*CMemosup5 1 27.4 16.76 16.76 <.0001 0.0003
victim3 1 1255 0.03 0.03 0.8578 0.8579
SMvictim3 1 1184 3.45 3.45 0.0631 0.0634
CMvictim3 1 32.7 0.20 0.20 0.6524 0.6554
time1*victim3 1 1504 0.00 0.00 0.9765 0.9766
time1*SMvictim3 1 1458 0.13 0.13 0.7178 0.7179
time1*CMvictim3 1 31.1 0.01 0.01 0.9296 0.9302


Ch 11a: Add Random Within-Class Vicitimization Effect across Classes
Using Variable-Centered Level-1 and Level-2 Victim Predictors

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 11
Columns in Z Per Subject 53
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4631.64339684  
1 2 4098.66512895 0.02041882
2 3 4095.41516373 .
3 2 4094.48791433 0.00016398
4 1 4094.40498666 0.00000326
5 1 4094.40344373 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03147 0.01859 1.69 0.0452
UN(2,1) ClassID 0.02122 0.006550 3.24 0.0012
UN(2,2) ClassID 0.003663 0.004447 0.82 0.2051
UN(3,1) ClassID 0.01409 0.01056 1.33 0.1822
UN(3,2) ClassID 0.01078 0.005536 1.95 0.0515
UN(3,3) ClassID 0.001051 0.009660 0.11 0.4567
UN(1,1) ClassID*StudentID 0.3883 0.04414 8.80 <.0001
UN(2,1) ClassID*StudentID -0.01002 0.02056 -0.49 0.6259
UN(2,2) ClassID*StudentID 0.04906 0.01637 3.00 0.0014
Residual   0.3267 0.01947 16.78 <.0001

Fit Statistics
-2 Res Log Likelihood 4094.4
AIC (smaller is better) 4114.4
AICC (smaller is better) 4114.5
BIC (smaller is better) 4129.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 537.24 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4094.4 10 4114.4 4114.5 4119.4 4129.4 4139.4

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2194 0.06129 59.5 68.84 <.0001 0.05 4.0968 4.3421
time1 -0.1306 0.02941 80.1 -4.44 <.0001 0.05 -0.1891 -0.07205
girl 0.2674 0.06839 553 3.91 0.0001 0.05 0.1331 0.4017
time1*girl -0.00780 0.03939 550 -0.20 0.8431 0.05 -0.08516 0.06957
CMgirl50 0.5617 0.4722 32.6 1.19 0.2428 0.05 -0.3995 1.5230
time1*CMgirl50 -1.0275 0.2133 27.3 -4.82 <.0001 0.05 -1.4649 -0.5901
CMemosup5 0.08658 0.07784 29.4 1.11 0.2750 0.05 -0.07252 0.2457
time1*CMemosup5 0.1879 0.03427 22.5 5.48 <.0001 0.05 0.1169 0.2589
WSvictim 0.01145 0.03067 1101 0.37 0.7089 0.05 -0.04872 0.07162
WCvictim -0.1243 0.03572 32 -3.48 0.0015 0.05 -0.1970 -0.05151
CMvictim3 -0.03256 0.1738 33.9 -0.19 0.8525 0.05 -0.3858 0.3206

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 80.1 19.71 19.71 <.0001 <.0001
girl 1 553 15.29 15.29 <.0001 0.0001
time1*girl 1 550 0.04 0.04 0.8431 0.8431
CMgirl50 1 32.6 1.41 1.41 0.2343 0.2428
time1*CMgirl50 1 27.3 23.21 23.21 <.0001 <.0001
CMemosup5 1 29.4 1.24 1.24 0.2660 0.2750
time1*CMemosup5 1 22.5 30.06 30.06 <.0001 <.0001
WSvictim 1 1101 0.14 0.14 0.7088 0.7089
WCvictim 1 32 12.11 12.11 0.0005 0.0015
CMvictim3 1 33.9 0.04 0.04 0.8514 0.8525


Likelihood Ratio Test for FitVicVBC vs. FitVicRWC3C

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitVicVBC 4099.0 7 4113.0 4123.5 . . .
FitVicRWC3C 4094.4 10 4114.4 4129.4 4.57526 3 0.20567


Ch 11a: Add Random Within-Student Vicitimization Effect across Students
Using Variable-Centered Level-1 and Level-2 Victim Predictors

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 11
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4631.64339684  
1 3 4083.20285214 0.00857765
2 2 4080.65014837 0.00101641
3 1 4080.17135297 0.00000743
4 1 4080.16791672 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03228 0.01875 1.72 0.0426
UN(2,1) ClassID 0.02103 0.006775 3.10 0.0019
UN(2,2) ClassID 0.004998 0.004747 1.05 0.1462
UN(1,1) ClassID*StudentID 0.3944 0.04295 9.18 <.0001
UN(2,1) ClassID*StudentID -0.00509 0.02008 -0.25 0.7999
UN(2,2) ClassID*StudentID 0.04456 0.01609 2.77 0.0028
UN(3,1) ClassID*StudentID 0.01991 0.02737 0.73 0.4670
UN(3,2) ClassID*StudentID -0.01261 0.01353 -0.93 0.3516
UN(3,3) ClassID*StudentID 0.1020 0.03225 3.16 0.0008
Residual   0.2957 0.01971 15.00 <.0001

Fit Statistics
-2 Res Log Likelihood 4080.2
AIC (smaller is better) 4100.2
AICC (smaller is better) 4100.3
BIC (smaller is better) 4115.1

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 551.48 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4080.2 10 4100.2 4100.3 4105.2 4115.1 4125.1

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2224 0.06148 58.6 68.68 <.0001 0.05 4.0994 4.3455
time1 -0.1296 0.02975 73.4 -4.36 <.0001 0.05 -0.1889 -0.07034
girl 0.2535 0.06826 558 3.71 0.0002 0.05 0.1195 0.3876
time1*girl -0.00171 0.03867 540 -0.04 0.9648 0.05 -0.07768 0.07427
CMgirl50 0.5455 0.4876 32.2 1.12 0.2715 0.05 -0.4475 1.5386
time1*CMgirl50 -1.0963 0.2372 31 -4.62 <.0001 0.05 -1.5800 -0.6125
CMemosup5 0.07119 0.08051 29.3 0.88 0.3838 0.05 -0.09340 0.2358
time1*CMemosup5 0.1661 0.03869 26.8 4.29 0.0002 0.05 0.08666 0.2455
WSvictim -0.01610 0.03600 278 -0.45 0.6551 0.05 -0.08698 0.05478
WCvictim -0.1300 0.03571 559 -3.64 0.0003 0.05 -0.2002 -0.05988
CMvictim3 -0.03159 0.1757 33.5 -0.18 0.8584 0.05 -0.3889 0.3257

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 73.4 18.98 18.98 <.0001 <.0001
girl 1 558 13.80 13.80 0.0002 0.0002
time1*girl 1 540 0.00 0.00 0.9648 0.9648
CMgirl50 1 32.2 1.25 1.25 0.2633 0.2715
time1*CMgirl50 1 31 21.36 21.36 <.0001 <.0001
CMemosup5 1 29.3 0.78 0.78 0.3766 0.3838
time1*CMemosup5 1 26.8 18.43 18.43 <.0001 0.0002
WSvictim 1 278 0.20 0.20 0.6548 0.6551
WCvictim 1 559 13.26 13.26 0.0003 0.0003
CMvictim3 1 33.5 0.03 0.03 0.8573 0.8584


Likelihood Ratio Test for FitVicVBC vs. FitVicRWS2C

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitVicVBC 4099.0 7 4113.0 4123.5 . . .
FitVicRWS2C 4080.2 10 4100.2 4115.1 18.8108 3 .000299165


Ch 11a: Add Random Within-Student Vicitimization Effect across Classes
Using Variable-Centered Level-1 and Level-2 Victim Predictors

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 13
Columns in X 11
Columns in Z Per Subject 78
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4631.64339684  
1 2 4080.16953475 0.01568818
2 3 4076.29816391 .
3 1 4075.50366483 0.00004846
4 1 4075.48063881 0.00000022
5 1 4075.48053965 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03241 0.01890 1.71 0.0432
UN(2,1) ClassID 0.02082 0.006603 3.15 0.0016
UN(2,2) ClassID 0.004210 0.004525 0.93 0.1761
UN(3,1) ClassID 0.01475 0.01071 1.38 0.1685
UN(3,2) ClassID 0.01085 0.005559 1.95 0.0508
UN(3,3) ClassID 0.000902 0.009672 0.09 0.4628
UN(1,1) ClassID*StudentID 0.4032 0.04432 9.10 <.0001
UN(2,1) ClassID*StudentID -0.00986 0.02033 -0.48 0.6279
UN(2,2) ClassID*StudentID 0.04518 0.01615 2.80 0.0026
UN(3,1) ClassID*StudentID 0.02784 0.02787 1.00 0.3179
UN(3,2) ClassID*StudentID -0.01393 0.01353 -1.03 0.3032
UN(3,3) ClassID*StudentID 0.1021 0.03244 3.15 0.0008
Residual   0.2956 0.01971 15.00 <.0001

Fit Statistics
-2 Res Log Likelihood 4075.5
AIC (smaller is better) 4101.5
AICC (smaller is better) 4101.7
BIC (smaller is better) 4120.9

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
12 556.16 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4075.5 13 4101.5 4101.7 4108.0 4120.9 4133.9

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2186 0.06154 58.5 68.55 <.0001 0.05 4.0955 4.3418
time1 -0.1296 0.02935 76.8 -4.42 <.0001 0.05 -0.1880 -0.07114
girl 0.2635 0.06829 554 3.86 0.0001 0.05 0.1294 0.3977
time1*girl -0.00250 0.03870 539 -0.06 0.9486 0.05 -0.07852 0.07352
CMgirl50 0.5972 0.4743 32.3 1.26 0.2171 0.05 -0.3686 1.5629
time1*CMgirl50 -1.0581 0.2143 27.1 -4.94 <.0001 0.05 -1.4977 -0.6185
CMemosup5 0.08913 0.07793 28.9 1.14 0.2621 0.05 -0.07027 0.2485
time1*CMemosup5 0.1833 0.03424 22.1 5.35 <.0001 0.05 0.1123 0.2543
WSvictim -0.01543 0.03595 274 -0.43 0.6681 0.05 -0.08621 0.05534
WCvictim -0.1261 0.03565 31.6 -3.54 0.0013 0.05 -0.1988 -0.05347
CMvictim3 -0.04074 0.1749 33.4 -0.23 0.8172 0.05 -0.3964 0.3150

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 76.8 19.50 19.50 <.0001 <.0001
girl 1 554 14.89 14.89 0.0001 0.0001
time1*girl 1 539 0.00 0.00 0.9485 0.9486
CMgirl50 1 32.3 1.58 1.58 0.2081 0.2171
time1*CMgirl50 1 27.1 24.38 24.38 <.0001 <.0001
CMemosup5 1 28.9 1.31 1.31 0.2527 0.2621
time1*CMemosup5 1 22.1 28.65 28.65 <.0001 <.0001
WSvictim 1 274 0.18 0.18 0.6677 0.6681
WCvictim 1 31.6 12.52 12.52 0.0004 0.0013
CMvictim3 1 33.4 0.05 0.05 0.8158 0.8172


Likelihood Ratio Test for FitVicRWS2C vs. FitVicRWS23C

Name Neg2LogLike Parms AIC BIC DevDiff DFdiff Pvalue
FitVicRWS2C 4080.2 10 4100.2 4115.1 . . .
FitVicRWS23C 4075.5 13 4101.5 4120.9 4.68738 3 0.19617


Ch 11a: Add Quadratic Level-3 Effects of Gender
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 13
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4617.05137949  
1 3 4075.63125191 0.00965330
2 2 4072.98239918 0.00120781
3 1 4072.41817978 0.00000995
4 1 4072.41359089 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03377 0.01959 1.72 0.0423
UN(2,1) ClassID 0.02078 0.006963 2.98 0.0028
UN(2,2) ClassID 0.004835 0.004874 0.99 0.1606
UN(1,1) ClassID*StudentID 0.3948 0.04298 9.18 <.0001
UN(2,1) ClassID*StudentID -0.00531 0.02009 -0.26 0.7917
UN(2,2) ClassID*StudentID 0.04474 0.01611 2.78 0.0027
UN(3,1) ClassID*StudentID 0.02029 0.02740 0.74 0.4590
UN(3,2) ClassID*StudentID -0.01228 0.01355 -0.91 0.3648
UN(3,3) ClassID*StudentID 0.1024 0.03227 3.17 0.0008
Residual   0.2956 0.01970 15.00 <.0001

Fit Statistics
-2 Res Log Likelihood 4072.4
AIC (smaller is better) 4092.4
AICC (smaller is better) 4092.5
BIC (smaller is better) 4107.4

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 544.64 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4072.4 10 4092.4 4092.5 4097.4 4107.4 4117.4

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2404 0.06834 45.6 62.05 <.0001 0.05 4.1028 4.3780
time1 -0.1162 0.03346 50.3 -3.47 0.0011 0.05 -0.1834 -0.04900
girl 0.2533 0.06827 557 3.71 0.0002 0.05 0.1192 0.3874
time1*girl -0.00152 0.03870 540 -0.04 0.9688 0.05 -0.07753 0.07450
CMgirl50 0.6432 0.5228 30.1 1.23 0.2281 0.05 -0.4243 1.7108
time1*CMgirl50 -1.0341 0.2471 28 -4.19 0.0003 0.05 -1.5402 -0.5280
CMemosup5 0.06595 0.08165 27.8 0.81 0.4261 0.05 -0.1014 0.2333
time1*CMemosup5 0.1624 0.03876 24.8 4.19 0.0003 0.05 0.08254 0.2423
WSvictim -0.01650 0.03605 279 -0.46 0.6474 0.05 -0.08746 0.05445
WCvictim -0.1300 0.03571 559 -3.64 0.0003 0.05 -0.2002 -0.05989
CMvictim3 -0.02289 0.1813 32.2 -0.13 0.9003 0.05 -0.3921 0.3464
CMgirl50*CMgirl50 -1.8169 3.2979 28.5 -0.55 0.5860 0.05 -8.5667 4.9329
time1*CMgirl50*CMgirl50 -1.3877 1.5999 26.7 -0.87 0.3935 0.05 -4.6720 1.8966

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 50.3 12.06 12.06 0.0005 0.0011
girl 1 557 13.77 13.77 0.0002 0.0002
time1*girl 1 540 0.00 0.00 0.9688 0.9688
CMgirl50 1 30.1 1.51 1.51 0.2185 0.2281
time1*CMgirl50 1 28 17.52 17.52 <.0001 0.0003
CMemosup5 1 27.8 0.65 0.65 0.4193 0.4261
time1*CMemosup5 1 24.8 17.55 17.55 <.0001 0.0003
WSvictim 1 279 0.21 0.21 0.6470 0.6474
WCvictim 1 559 13.26 13.26 0.0003 0.0003
CMvictim3 1 32.2 0.02 0.02 0.8995 0.9003
CMgirl50*CMgirl50 1 28.5 0.30 0.30 0.5817 0.5860
time1*CMgirl50*CMgirl50 1 26.7 0.75 0.75 0.3857 0.3935


Ch 11a: Add Quadratic Level-3 Effects of Emotional Support
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 13
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4636.97652093  
1 3 4089.32135552 0.00906185
2 2 4086.47587460 0.00102396
3 1 4085.98004471 0.00001015
4 1 4085.97528466 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.03363 0.01940 1.73 0.0415
UN(2,1) ClassID 0.02125 0.007049 3.01 0.0026
UN(2,2) ClassID 0.005492 0.004968 1.11 0.1345
UN(1,1) ClassID*StudentID 0.3945 0.04296 9.18 <.0001
UN(2,1) ClassID*StudentID -0.00516 0.02009 -0.26 0.7973
UN(2,2) ClassID*StudentID 0.04458 0.01610 2.77 0.0028
UN(3,1) ClassID*StudentID 0.02022 0.02740 0.74 0.4605
UN(3,2) ClassID*StudentID -0.01260 0.01355 -0.93 0.3525
UN(3,3) ClassID*StudentID 0.1022 0.03227 3.17 0.0008
Residual   0.2957 0.01971 15.00 <.0001

Fit Statistics
-2 Res Log Likelihood 4086.0
AIC (smaller is better) 4106.0
AICC (smaller is better) 4106.1
BIC (smaller is better) 4120.9

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 551.00 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4086.0 10 4106.0 4106.1 4111.0 4120.9 4130.9

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2573 0.08242 40.1 51.65 <.0001 0.05 4.0907 4.4239
time1 -0.1210 0.03662 47.1 -3.31 0.0018 0.05 -0.1947 -0.04736
girl 0.2535 0.06826 558 3.71 0.0002 0.05 0.1195 0.3876
time1*girl -0.00168 0.03868 540 -0.04 0.9655 0.05 -0.07766 0.07431
CMgirl50 0.5301 0.4932 31.2 1.07 0.2907 0.05 -0.4756 1.5357
time1*CMgirl50 -1.0920 0.2407 29.9 -4.54 <.0001 0.05 -1.5835 -0.6004
CMemosup5 0.03339 0.1004 28.9 0.33 0.7418 0.05 -0.1719 0.2387
time1*CMemosup5 0.1568 0.04552 25.4 3.45 0.0020 0.05 0.06316 0.2505
WSvictim -0.01644 0.03604 278 -0.46 0.6487 0.05 -0.08737 0.05450
WCvictim -0.1301 0.03571 559 -3.64 0.0003 0.05 -0.2002 -0.05991
CMvictim3 -0.07326 0.1924 31.7 -0.38 0.7060 0.05 -0.4654 0.3188
CMemosup5*CMemosup5 -0.08272 0.1269 26.8 -0.65 0.5201 0.05 -0.3432 0.1778
time1*CMemosup5*CMemosup5 -0.02354 0.05822 24 -0.40 0.6896 0.05 -0.1437 0.09661

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 47.1 10.92 10.92 0.0009 0.0018
girl 1 558 13.80 13.80 0.0002 0.0002
time1*girl 1 540 0.00 0.00 0.9654 0.9655
CMgirl50 1 31.2 1.16 1.16 0.2825 0.2907
time1*CMgirl50 1 29.9 20.59 20.59 <.0001 <.0001
CMemosup5 1 28.9 0.11 0.11 0.7394 0.7418
time1*CMemosup5 1 25.4 11.87 11.87 0.0006 0.0020
WSvictim 1 278 0.21 0.21 0.6483 0.6487
WCvictim 1 559 13.26 13.26 0.0003 0.0003
CMvictim3 1 31.7 0.14 0.14 0.7034 0.7060
CMemosup5*CMemosup5 1 26.8 0.42 0.42 0.5145 0.5201
time1*CMemosup5*CMemosup5 1 24 0.16 0.16 0.6860 0.6896


Ch 11a: Add Quadratic Effect of Victimization at Each Level
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 14
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4618.98890536  
1 3 4087.31171511 0.00819167
2 2 4084.66253645 0.00114989
3 1 4084.09558265 0.00001498
4 1 4084.08855090 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02825 0.01802 1.57 0.0585
UN(2,1) ClassID 0.02009 0.006682 3.01 0.0026
UN(2,2) ClassID 0.005423 0.004874 1.11 0.1329
UN(1,1) ClassID*StudentID 0.3938 0.04310 9.14 <.0001
UN(2,1) ClassID*StudentID -0.00466 0.02010 -0.23 0.8165
UN(2,2) ClassID*StudentID 0.04251 0.01609 2.64 0.0041
UN(3,1) ClassID*StudentID 0.02131 0.02736 0.78 0.4360
UN(3,2) ClassID*StudentID -0.01239 0.01347 -0.92 0.3576
UN(3,3) ClassID*StudentID 0.09630 0.03209 3.00 0.0013
Residual   0.2985 0.01999 14.93 <.0001

Fit Statistics
-2 Res Log Likelihood 4084.1
AIC (smaller is better) 4104.1
AICC (smaller is better) 4104.2
BIC (smaller is better) 4119.1

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 534.90 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4084.1 10 4104.1 4104.2 4109.1 4119.1 4129.1

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.1970 0.07126 68.4 58.90 <.0001 0.05 4.0549 4.3392
time1 -0.1305 0.02991 71.5 -4.36 <.0001 0.05 -0.1901 -0.07089
girl 0.2505 0.06834 557 3.67 0.0003 0.05 0.1162 0.3847
time1*girl -0.00093 0.03857 541 -0.02 0.9808 0.05 -0.07669 0.07484
CMgirl50 0.5442 0.4752 31.2 1.15 0.2609 0.05 -0.4247 1.5131
time1*CMgirl50 -1.0847 0.2395 30.6 -4.53 <.0001 0.05 -1.5735 -0.5960
CMemosup5 0.07948 0.07856 28.3 1.01 0.3202 0.05 -0.08136 0.2403
time1*CMemosup5 0.1648 0.03911 26.5 4.21 0.0003 0.05 0.08451 0.2451
WSvictim -0.01519 0.03578 269 -0.42 0.6716 0.05 -0.08564 0.05527
WCvictim -0.1291 0.04244 588 -3.04 0.0024 0.05 -0.2125 -0.04578
CMvictim3 -0.1520 0.1982 29.3 -0.77 0.4493 0.05 -0.5573 0.2533
CMvictim3*CMvictim3 0.5862 0.4475 29.7 1.31 0.2002 0.05 -0.3281 1.5005
WCvictim*WCvictim 0.01420 0.03395 581 0.42 0.6760 0.05 -0.05249 0.08088
WSvictim*WSvictim -0.08793 0.04966 1027 -1.77 0.0770 0.05 -0.1854 0.009528

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 71.5 19.04 19.04 <.0001 <.0001
girl 1 557 13.43 13.43 0.0002 0.0003
time1*girl 1 541 0.00 0.00 0.9808 0.9808
CMgirl50 1 31.2 1.31 1.31 0.2522 0.2609
time1*CMgirl50 1 30.6 20.51 20.51 <.0001 <.0001
CMemosup5 1 28.3 1.02 1.02 0.3117 0.3202
time1*CMemosup5 1 26.5 17.76 17.76 <.0001 0.0003
WSvictim 1 269 0.18 0.18 0.6713 0.6716
WCvictim 1 588 9.26 9.26 0.0023 0.0024
CMvictim3 1 29.3 0.59 0.59 0.4432 0.4493
CMvictim3*CMvictim3 1 29.7 1.72 1.72 0.1902 0.2002
WCvictim*WCvictim 1 581 0.17 0.17 0.6759 0.6760
WSvictim*WSvictim 1 1027 3.13 3.13 0.0767 0.0770


Ch 11a: Add Two-Way and Three-Way Interactions Among Level-3 Effects
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 19
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4592.72612570  
1 3 4058.24044742 0.00652556
2 2 4055.92563513 0.00044981
3 1 4055.71319303 0.00000257
4 1 4055.71201691 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02249 0.01668 1.35 0.0888
UN(2,1) ClassID 0.02132 0.006778 3.14 0.0017
UN(2,2) ClassID 0.006048 0.005186 1.17 0.1218
UN(1,1) ClassID*StudentID 0.3937 0.04299 9.16 <.0001
UN(2,1) ClassID*StudentID -0.00558 0.02013 -0.28 0.7817
UN(2,2) ClassID*StudentID 0.04490 0.01613 2.78 0.0027
UN(3,1) ClassID*StudentID 0.02808 0.02769 1.01 0.3106
UN(3,2) ClassID*StudentID -0.01550 0.01380 -1.12 0.2614
UN(3,3) ClassID*StudentID 0.1003 0.03220 3.12 0.0009
Residual   0.2962 0.01977 14.98 <.0001

Fit Statistics
-2 Res Log Likelihood 4055.7
AIC (smaller is better) 4075.7
AICC (smaller is better) 4075.8
BIC (smaller is better) 4090.7

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 537.01 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4055.7 10 4075.7 4075.8 4080.7 4090.7 4100.7

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2040 0.06273 53.1 67.02 <.0001 0.05 4.0782 4.3298
time1 -0.1296 0.03230 57.6 -4.01 0.0002 0.05 -0.1942 -0.06490
girl 0.2551 0.06816 558 3.74 0.0002 0.05 0.1212 0.3890
time1*girl -0.00267 0.03867 540 -0.07 0.9450 0.05 -0.07864 0.07330
CMgirl50 0.6115 0.5887 32.3 1.04 0.3067 0.05 -0.5873 1.8102
time1*CMgirl50 -0.9913 0.3141 31.3 -3.16 0.0035 0.05 -1.6317 -0.3509
CMemosup5 0.1858 0.1124 26.3 1.65 0.1102 0.05 -0.04511 0.4168
time1*CMemosup5 0.1708 0.05969 24.4 2.86 0.0085 0.05 0.04770 0.2939
WSvictim -0.01665 0.03598 276 -0.46 0.6439 0.05 -0.08748 0.05418
WCvictim -0.1312 0.03567 560 -3.68 0.0003 0.05 -0.2012 -0.06112
CMvictim3 -0.04077 0.1666 31.7 -0.24 0.8082 0.05 -0.3801 0.2986
CMgirl50*CMemosup5 -2.9451 0.9968 32.5 -2.95 0.0058 0.05 -4.9744 -0.9158
CMgirl50*CMvictim3 -0.09553 2.1076 31.9 -0.05 0.9641 0.05 -4.3893 4.1982
CMemosup5*CMvictim3 -0.6954 0.3041 26.2 -2.29 0.0305 0.05 -1.3201 -0.07057
time1*CMgirl50*CMemosup5 -0.4601 0.5605 33 -0.82 0.4177 0.05 -1.6005 0.6804
time1*CMgirl50*CMvictim3 -0.8900 1.1610 31.4 -0.77 0.4491 0.05 -3.2567 1.4768
time1*CMemosup5*CMvictim3 0.08136 0.1672 25.2 0.49 0.6307 0.05 -0.2628 0.4256
CMgirl50*CMemosup5*CMvictim3 4.2891 4.2895 27.6 1.00 0.3260 0.05 -4.5027 13.0809
time1*CMgirl50*CMemosup5*CMvictim3 0.4091 2.3611 27.1 0.17 0.8637 0.05 -4.4347 5.2530

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 57.6 16.09 16.09 <.0001 0.0002
girl 1 558 14.01 14.01 0.0002 0.0002
time1*girl 1 540 0.00 0.00 0.9450 0.9450
CMgirl50 1 32.3 1.08 1.08 0.2990 0.3067
time1*CMgirl50 1 31.3 9.96 9.96 0.0016 0.0035
CMemosup5 1 26.3 2.73 2.73 0.0983 0.1102
time1*CMemosup5 1 24.4 8.19 8.19 0.0042 0.0085
WSvictim 1 276 0.21 0.21 0.6435 0.6439
WCvictim 1 560 13.53 13.53 0.0002 0.0003
CMvictim3 1 31.7 0.06 0.06 0.8066 0.8082
CMgirl50*CMemosup5 1 32.5 8.73 8.73 0.0031 0.0058
CMgirl50*CMvictim3 1 31.9 0.00 0.00 0.9638 0.9641
CMemosup5*CMvictim3 1 26.2 5.23 5.23 0.0222 0.0305
time1*CMgirl50*CMemosup5 1 33 0.67 0.67 0.4118 0.4177
time1*CMgirl50*CMvictim3 1 31.4 0.59 0.59 0.4433 0.4491
time1*CMemosup5*CMvictim3 1 25.2 0.24 0.24 0.6265 0.6307
CMgirl50*CMemosup5*CMvictim3 1 27.6 1.00 1.00 0.3173 0.3260
time1*CMgirl50*CMemosup5*CMvictim3 1 27.1 0.03 0.03 0.8624 0.8637


Ch 11a: Keep Gender*Emotional Support and Emotional Support*Victimization Level-3 Interactions
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 13
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4607.40112491  
1 3 4074.05997749 0.00875499
2 2 4071.55615495 0.00117896
3 2 4071.01608709 0.00000985
4 1 4071.01155021 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02135 0.01580 1.35 0.0883
UN(2,1) ClassID 0.02018 0.006172 3.27 0.0011
UN(2,2) ClassID 0.004635 0.004571 1.01 0.1553
UN(1,1) ClassID*StudentID 0.3925 0.04286 9.16 <.0001
UN(2,1) ClassID*StudentID -0.00495 0.02007 -0.25 0.8050
UN(2,2) ClassID*StudentID 0.04441 0.01611 2.76 0.0029
UN(3,1) ClassID*StudentID 0.02437 0.02749 0.89 0.3754
UN(3,2) ClassID*StudentID -0.01473 0.01357 -1.09 0.2777
UN(3,3) ClassID*StudentID 0.1018 0.03234 3.15 0.0008
Residual   0.2963 0.01977 14.99 <.0001

Fit Statistics
-2 Res Log Likelihood 4071.0
AIC (smaller is better) 4091.0
AICC (smaller is better) 4091.1
BIC (smaller is better) 4106.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 536.39 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4071.0 10 4091.0 4091.1 4096.0 4106.0 4116.0

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2122 0.05863 64.5 71.84 <.0001 0.05 4.0951 4.3293
time1 -0.1302 0.02956 76.9 -4.40 <.0001 0.05 -0.1890 -0.07133
girl 0.2536 0.06814 560 3.72 0.0002 0.05 0.1198 0.3874
time1*girl -0.00190 0.03866 541 -0.05 0.9608 0.05 -0.07783 0.07403
CMgirl50 0.5357 0.4508 32.5 1.19 0.2433 0.05 -0.3820 1.4534
time1*CMgirl50 -1.1193 0.2349 32 -4.77 <.0001 0.05 -1.5977 -0.6408
CMemosup5 0.1051 0.07544 29.6 1.39 0.1741 0.05 -0.04909 0.2592
time1*CMemosup5 0.1694 0.03828 27.6 4.42 0.0001 0.05 0.09090 0.2478
WSvictim -0.01824 0.03596 275 -0.51 0.6124 0.05 -0.08904 0.05256
WCvictim -0.1303 0.03565 561 -3.65 0.0003 0.05 -0.2003 -0.06027
CMvictim3 -0.06424 0.1620 34.3 -0.40 0.6942 0.05 -0.3934 0.2649
CMgirl50*CMemosup5 -2.3264 0.8212 37.5 -2.83 0.0074 0.05 -3.9895 -0.6633
CMemosup5*CMvictim3 -0.7479 0.2832 31.9 -2.64 0.0127 0.05 -1.3248 -0.1710

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 76.9 19.40 19.40 <.0001 <.0001
girl 1 560 13.85 13.85 0.0002 0.0002
time1*girl 1 541 0.00 0.00 0.9608 0.9608
CMgirl50 1 32.5 1.41 1.41 0.2347 0.2433
time1*CMgirl50 1 32 22.71 22.71 <.0001 <.0001
CMemosup5 1 29.6 1.94 1.94 0.1637 0.1741
time1*CMemosup5 1 27.6 19.58 19.58 <.0001 0.0001
WSvictim 1 275 0.26 0.26 0.6120 0.6124
WCvictim 1 561 13.36 13.36 0.0003 0.0003
CMvictim3 1 34.3 0.16 0.16 0.6917 0.6942
CMgirl50*CMemosup5 1 37.5 8.02 8.02 0.0046 0.0074
CMemosup5*CMvictim3 1 31.9 6.97 6.97 0.0083 0.0127


PsuedoR2 (% Reduction) for CovRandVic vs. CovL3x

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
CovRandVic UN(1,1) ClassID 0.03228 0.01875 1.72 0.0426 .
CovRandVic UN(2,2) ClassID 0.004998 0.004747 1.05 0.1462 .
CovRandVic UN(1,1) ClassID*StudentID 0.3944 0.04295 9.18 <.0001 .
CovRandVic UN(2,2) ClassID*StudentID 0.04456 0.01609 2.77 0.0028 .
CovRandVic UN(3,3) ClassID*StudentID 0.1020 0.03225 3.16 0.0008 .
CovRandVic Residual   0.2957 0.01971 15.00 <.0001 .
CovL3x UN(1,1) ClassID 0.02135 0.01580 1.35 0.0883 0.33863
CovL3x UN(2,2) ClassID 0.004635 0.004571 1.01 0.1553 0.07260
CovL3x UN(1,1) ClassID*StudentID 0.3925 0.04286 9.16 <.0001 0.00474
CovL3x UN(2,2) ClassID*StudentID 0.04441 0.01611 2.76 0.0029 0.00323
CovL3x UN(3,3) ClassID*StudentID 0.1018 0.03234 3.15 0.0008 0.00188
CovL3x Residual   0.2963 0.01977 14.99 <.0001 -0.00229


Eq 11a.8: Add Level-2 Interactions (and Contextual Level-3 Interactions)
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 15
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4601.42414688  
1 3 4072.97056674 0.00882650
2 2 4070.43992816 0.00118475
3 2 4069.89783492 0.00001049
4 1 4069.89298694 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.01997 0.01542 1.30 0.0976
UN(2,1) ClassID 0.01991 0.006068 3.28 0.0010
UN(2,2) ClassID 0.004571 0.004551 1.00 0.1576
UN(1,1) ClassID*StudentID 0.3945 0.04308 9.16 <.0001
UN(2,1) ClassID*StudentID -0.00718 0.02015 -0.36 0.7214
UN(2,2) ClassID*StudentID 0.04425 0.01610 2.75 0.0030
UN(3,1) ClassID*StudentID 0.02451 0.02758 0.89 0.3741
UN(3,2) ClassID*StudentID -0.01459 0.01357 -1.08 0.2822
UN(3,3) ClassID*StudentID 0.1023 0.03241 3.16 0.0008
Residual   0.2963 0.01977 14.99 <.0001

Fit Statistics
-2 Res Log Likelihood 4069.9
AIC (smaller is better) 4089.9
AICC (smaller is better) 4090.0
BIC (smaller is better) 4104.9

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 531.53 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4069.9 10 4089.9 4090.0 4094.9 4104.9 4114.9

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2137 0.05827 65.8 72.31 <.0001 0.05 4.0973 4.3300
time1 -0.1305 0.02952 77.2 -4.42 <.0001 0.05 -0.1893 -0.07176
girl 0.2530 0.06826 558 3.71 0.0002 0.05 0.1190 0.3871
time1*girl -0.00139 0.03865 541 -0.04 0.9713 0.05 -0.07731 0.07452
CMgirl50 0.5432 0.4464 32.7 1.22 0.2323 0.05 -0.3652 1.4517
time1*CMgirl50 -1.1189 0.2344 32 -4.77 <.0001 0.05 -1.5964 -0.6414
CMemosup5 0.1055 0.07466 29.8 1.41 0.1682 0.05 -0.04706 0.2580
time1*CMemosup5 0.1695 0.03820 27.6 4.44 0.0001 0.05 0.09125 0.2478
WSvictim -0.01806 0.03599 276 -0.50 0.6162 0.05 -0.08892 0.05279
WCvictim -0.05512 0.05166 572 -1.07 0.2864 0.05 -0.1566 0.04634
CMvictim3 -0.07057 0.1608 34.6 -0.44 0.6635 0.05 -0.3972 0.2560
CMgirl50*CMemosup5 -2.3128 0.8151 37.8 -2.84 0.0073 0.05 -3.9632 -0.6625
CMemosup5*CMvictim3 -0.7586 0.2811 32.2 -2.70 0.0110 0.05 -1.3310 -0.1862
girl*WCvictim -0.1527 0.07306 581 -2.09 0.0371 0.05 -0.2961 -0.00918
CMgirl50*WCvictim -0.09640 0.3708 561 -0.26 0.7950 0.05 -0.8247 0.6319

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 77.2 19.56 19.56 <.0001 <.0001
girl 1 558 13.74 13.74 0.0002 0.0002
time1*girl 1 541 0.00 0.00 0.9713 0.9713
CMgirl50 1 32.7 1.48 1.48 0.2236 0.2323
time1*CMgirl50 1 32 22.78 22.78 <.0001 <.0001
CMemosup5 1 29.8 1.99 1.99 0.1578 0.1682
time1*CMemosup5 1 27.6 19.70 19.70 <.0001 0.0001
WSvictim 1 276 0.25 0.25 0.6158 0.6162
WCvictim 1 572 1.14 1.14 0.2860 0.2864
CMvictim3 1 34.6 0.19 0.19 0.6608 0.6635
CMgirl50*CMemosup5 1 37.8 8.05 8.05 0.0045 0.0073
CMemosup5*CMvictim3 1 32.2 7.28 7.28 0.0070 0.0110
girl*WCvictim 1 581 4.37 4.37 0.0366 0.0371
CMgirl50*WCvictim 1 561 0.07 0.07 0.7949 0.7950


PsuedoR2 (% Reduction) for CovL3x vs. CovL32x

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
CovL3x UN(1,1) ClassID 0.02135 0.01580 1.35 0.0883 .
CovL3x UN(2,2) ClassID 0.004635 0.004571 1.01 0.1553 .
CovL3x UN(1,1) ClassID*StudentID 0.3925 0.04286 9.16 <.0001 .
CovL3x UN(2,2) ClassID*StudentID 0.04441 0.01611 2.76 0.0029 .
CovL3x UN(3,3) ClassID*StudentID 0.1018 0.03234 3.15 0.0008 .
CovL3x Residual   0.2963 0.01977 14.99 <.0001 .
CovL32x UN(1,1) ClassID 0.01997 0.01542 1.30 0.0976 0.064537
CovL32x UN(2,2) ClassID 0.004571 0.004551 1.00 0.1576 0.013785
CovL32x UN(1,1) ClassID*StudentID 0.3945 0.04308 9.16 <.0001 -0.004990
CovL32x UN(2,2) ClassID*StudentID 0.04425 0.01610 2.75 0.0030 0.003610
CovL32x UN(3,3) ClassID*StudentID 0.1023 0.03241 3.16 0.0008 -0.004537
CovL32x Residual   0.2963 0.01977 14.99 <.0001 0.000146


Ch 11a: Add Time by Level-2 Interactions (and Contextual Level-3 Interactions)
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 19
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4611.33974275  
1 3 4082.15901670 0.00825016
2 2 4079.54276837 0.00094147
3 2 4079.09794864 0.00000655
4 1 4079.09487296 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02054 0.01564 1.31 0.0946
UN(2,1) ClassID 0.01997 0.006255 3.19 0.0014
UN(2,2) ClassID 0.004955 0.004728 1.05 0.1473
UN(1,1) ClassID*StudentID 0.3941 0.04298 9.17 <.0001
UN(2,1) ClassID*StudentID -0.00633 0.02009 -0.31 0.7528
UN(2,2) ClassID*StudentID 0.04393 0.01606 2.74 0.0031
UN(3,1) ClassID*StudentID 0.02575 0.02770 0.93 0.3526
UN(3,2) ClassID*StudentID -0.01350 0.01369 -0.99 0.3241
UN(3,3) ClassID*StudentID 0.1052 0.03296 3.19 0.0007
Residual   0.2951 0.01974 14.95 <.0001

Fit Statistics
-2 Res Log Likelihood 4079.1
AIC (smaller is better) 4099.1
AICC (smaller is better) 4099.2
BIC (smaller is better) 4114.1

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 532.24 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4079.1 10 4099.1 4099.2 4104.1 4114.1 4124.1

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2131 0.05889 60.8 71.54 <.0001 0.05 4.0953 4.3309
time1 -0.1285 0.03187 60.4 -4.03 0.0002 0.05 -0.1922 -0.06475
girl 0.2546 0.06821 558 3.73 0.0002 0.05 0.1206 0.3886
time1*girl -0.00229 0.03864 539 -0.06 0.9527 0.05 -0.07820 0.07362
CMgirl50 0.5427 0.4496 31.8 1.21 0.2363 0.05 -0.3733 1.4588
time1*CMgirl50 -1.1205 0.2452 30.8 -4.57 <.0001 0.05 -1.6207 -0.6202
CMemosup5 0.1056 0.07516 29.1 1.40 0.1707 0.05 -0.04814 0.2593
time1*CMemosup5 0.1672 0.04003 27.1 4.18 0.0003 0.05 0.08507 0.2493
WSvictim -0.01872 0.03614 275 -0.52 0.6049 0.05 -0.08986 0.05242
WCvictim -0.07223 0.05794 578 -1.25 0.2130 0.05 -0.1860 0.04156
CMvictim3 -0.06831 0.1692 29.6 -0.40 0.6893 0.05 -0.4141 0.2774
CMgirl50*CMemosup5 -2.3182 0.8185 37.4 -2.83 0.0074 0.05 -3.9761 -0.6603
CMemosup5*CMvictim3 -0.7443 0.2825 31.8 -2.63 0.0129 0.05 -1.3198 -0.1687
girl*WCvictim -0.08575 0.08144 579 -1.05 0.2928 0.05 -0.2457 0.07421
CMgirl50*WCvictim 0.001601 0.4196 564 0.00 0.9970 0.05 -0.8226 0.8258
time1*WCvictim 0.02288 0.03284 561 0.70 0.4863 0.05 -0.04163 0.08739
time1*girl*WCvictim -0.08649 0.04642 567 -1.86 0.0630 0.05 -0.1777 0.004691
time1*CMvictim3 -0.00888 0.09102 27.9 -0.10 0.9230 0.05 -0.1954 0.1776
time1*CMgirl50*WCvictim -0.1261 0.2442 571 -0.52 0.6057 0.05 -0.6058 0.3536

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 60.4 16.25 16.25 <.0001 0.0002
girl 1 558 13.93 13.93 0.0002 0.0002
time1*girl 1 539 0.00 0.00 0.9527 0.9527
CMgirl50 1 31.8 1.46 1.46 0.2274 0.2363
time1*CMgirl50 1 30.8 20.88 20.88 <.0001 <.0001
CMemosup5 1 29.1 1.97 1.97 0.1601 0.1707
time1*CMemosup5 1 27.1 17.45 17.45 <.0001 0.0003
WSvictim 1 275 0.27 0.27 0.6045 0.6049
WCvictim 1 578 1.55 1.55 0.2125 0.2130
CMvictim3 1 29.6 0.16 0.16 0.6864 0.6893
CMgirl50*CMemosup5 1 37.4 8.02 8.02 0.0046 0.0074
CMemosup5*CMvictim3 1 31.8 6.94 6.94 0.0084 0.0129
girl*WCvictim 1 579 1.11 1.11 0.2924 0.2928
CMgirl50*WCvictim 1 564 0.00 0.00 0.9970 0.9970
time1*WCvictim 1 561 0.49 0.49 0.4860 0.4863
time1*girl*WCvictim 1 567 3.47 3.47 0.0624 0.0630
time1*CMvictim3 1 27.9 0.01 0.01 0.9223 0.9230
time1*CMgirl50*WCvictim 1 571 0.27 0.27 0.6055 0.6057


Ch 11a: Add Level-3 by Student Gender Cross-Level Interactions
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 18
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4602.04809008  
1 3 4075.22635987 0.00910745
2 2 4072.59569919 0.00129377
3 2 4071.99496458 0.00001228
4 1 4071.98923655 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.01991 0.01546 1.29 0.0989
UN(2,1) ClassID 0.01983 0.006078 3.26 0.0011
UN(2,2) ClassID 0.004583 0.004560 1.00 0.1575
UN(1,1) ClassID*StudentID 0.3975 0.04334 9.17 <.0001
UN(2,1) ClassID*StudentID -0.00780 0.02022 -0.39 0.6997
UN(2,2) ClassID*StudentID 0.04429 0.01610 2.75 0.0030
UN(3,1) ClassID*StudentID 0.02466 0.02774 0.89 0.3739
UN(3,2) ClassID*StudentID -0.01461 0.01358 -1.08 0.2820
UN(3,3) ClassID*StudentID 0.1024 0.03243 3.16 0.0008
Residual   0.2963 0.01977 14.99 <.0001

Fit Statistics
-2 Res Log Likelihood 4072.0
AIC (smaller is better) 4092.0
AICC (smaller is better) 4092.1
BIC (smaller is better) 4107.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 530.06 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4072.0 10 4092.0 4092.1 4097.0 4107.0 4117.0

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2165 0.06089 76.5 69.24 <.0001 0.05 4.0953 4.3378
time1 -0.1305 0.02953 77 -4.42 <.0001 0.05 -0.1893 -0.07172
girl 0.2528 0.07494 592 3.37 0.0008 0.05 0.1056 0.4000
time1*girl -0.00115 0.03865 541 -0.03 0.9763 0.05 -0.07708 0.07478
CMgirl50 0.7308 0.5867 87.3 1.25 0.2163 0.05 -0.4354 1.8969
time1*CMgirl50 -1.1188 0.2345 31.9 -4.77 <.0001 0.05 -1.5966 -0.6410
CMemosup5 0.1311 0.09270 68.5 1.41 0.1619 0.05 -0.05387 0.3160
time1*CMemosup5 0.1694 0.03822 27.6 4.43 0.0001 0.05 0.09103 0.2477
WSvictim -0.01799 0.03601 276 -0.50 0.6177 0.05 -0.08887 0.05289
WCvictim -0.05495 0.05177 569 -1.06 0.2890 0.05 -0.1566 0.04674
CMvictim3 -0.06482 0.2056 87.3 -0.32 0.7533 0.05 -0.4734 0.3438
CMgirl50*CMemosup5 -2.2301 0.8246 39.1 -2.70 0.0101 0.05 -3.8979 -0.5623
CMemosup5*CMvictim3 -0.7508 0.2821 32.2 -2.66 0.0120 0.05 -1.3253 -0.1764
girl*WCvictim -0.1527 0.07324 578 -2.08 0.0376 0.05 -0.2965 -0.00882
CMgirl50*WCvictim -0.1082 0.3743 559 -0.29 0.7725 0.05 -0.8434 0.6269
girl*CMgirl50 -0.3366 0.6742 578 -0.50 0.6178 0.05 -1.6608 0.9875
girl*CMemosup5 -0.05286 0.1084 557 -0.49 0.6260 0.05 -0.2658 0.1601
girl*CMvictim3 -0.00370 0.2449 564 -0.02 0.9880 0.05 -0.4847 0.4773

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 77 19.54 19.54 <.0001 <.0001
girl 1 592 11.38 11.38 0.0007 0.0008
time1*girl 1 541 0.00 0.00 0.9763 0.9763
CMgirl50 1 87.3 1.55 1.55 0.2130 0.2163
time1*CMgirl50 1 31.9 22.75 22.75 <.0001 <.0001
CMemosup5 1 68.5 2.00 2.00 0.1574 0.1619
time1*CMemosup5 1 27.6 19.64 19.64 <.0001 0.0001
WSvictim 1 276 0.25 0.25 0.6173 0.6177
WCvictim 1 569 1.13 1.13 0.2885 0.2890
CMvictim3 1 87.3 0.10 0.10 0.7526 0.7533
CMgirl50*CMemosup5 1 39.1 7.31 7.31 0.0068 0.0101
CMemosup5*CMvictim3 1 32.2 7.09 7.09 0.0078 0.0120
girl*WCvictim 1 578 4.35 4.35 0.0371 0.0376
CMgirl50*WCvictim 1 559 0.08 0.08 0.7724 0.7725
girl*CMgirl50 1 578 0.25 0.25 0.6176 0.6178
girl*CMemosup5 1 557 0.24 0.24 0.6258 0.6260
girl*CMvictim3 1 564 0.00 0.00 0.9880 0.9880


Ch 11a: Add Level-3 by Within-Class Victim Cross-Level Interactions
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 17
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4605.56723759  
1 3 4077.04844288 0.00884855
2 2 4074.50673200 0.00123811
3 2 4073.93428657 0.00001165
4 1 4073.92885092 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.01999 0.01541 1.30 0.0973
UN(2,1) ClassID 0.01984 0.006053 3.28 0.0010
UN(2,2) ClassID 0.004503 0.004530 0.99 0.1601
UN(1,1) ClassID*StudentID 0.3938 0.04307 9.14 <.0001
UN(2,1) ClassID*StudentID -0.00648 0.02014 -0.32 0.7477
UN(2,2) ClassID*StudentID 0.04416 0.01609 2.74 0.0030
UN(3,1) ClassID*StudentID 0.02352 0.02764 0.85 0.3949
UN(3,2) ClassID*StudentID -0.01459 0.01357 -1.07 0.2825
UN(3,3) ClassID*StudentID 0.1025 0.03246 3.16 0.0008
Residual   0.2963 0.01977 14.99 <.0001

Fit Statistics
-2 Res Log Likelihood 4073.9
AIC (smaller is better) 4093.9
AICC (smaller is better) 4094.1
BIC (smaller is better) 4108.9

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 531.64 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4073.9 10 4093.9 4094.1 4099.0 4108.9 4118.9

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2134 0.05826 65.8 72.32 <.0001 0.05 4.0971 4.3297
time1 -0.1306 0.02948 77.5 -4.43 <.0001 0.05 -0.1893 -0.07192
girl 0.2536 0.06825 557 3.72 0.0002 0.05 0.1196 0.3877
time1*girl -0.00123 0.03864 541 -0.03 0.9746 0.05 -0.07714 0.07467
CMgirl50 0.5452 0.4464 32.8 1.22 0.2307 0.05 -0.3632 1.4535
time1*CMgirl50 -1.1219 0.2340 32.1 -4.79 <.0001 0.05 -1.5985 -0.6453
CMemosup5 0.1054 0.07466 29.8 1.41 0.1683 0.05 -0.04710 0.2579
time1*CMemosup5 0.1698 0.03812 27.6 4.45 0.0001 0.05 0.09164 0.2479
WSvictim -0.01813 0.03600 275 -0.50 0.6150 0.05 -0.08899 0.05274
WCvictim -0.06492 0.05689 567 -1.14 0.2543 0.05 -0.1767 0.04683
CMvictim3 -0.07120 0.1608 34.6 -0.44 0.6607 0.05 -0.3978 0.2554
CMgirl50*CMemosup5 -2.3085 0.8151 37.8 -2.83 0.0074 0.05 -3.9589 -0.6581
CMemosup5*CMvictim3 -0.7581 0.2811 32.2 -2.70 0.0110 0.05 -1.3306 -0.1857
girl*WCvictim -0.1632 0.07358 580 -2.22 0.0269 0.05 -0.3077 -0.01869
CMgirl50*WCvictim 0.05491 0.3947 559 0.14 0.8894 0.05 -0.7204 0.8302
CMemosup5*WCvictim -0.05734 0.06299 561 -0.91 0.3630 0.05 -0.1811 0.06639
WCvictim*CMvictim3 0.08979 0.1590 552 0.56 0.5726 0.05 -0.2226 0.4022

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 77.5 19.63 19.63 <.0001 <.0001
girl 1 557 13.81 13.81 0.0002 0.0002
time1*girl 1 541 0.00 0.00 0.9746 0.9746
CMgirl50 1 32.8 1.49 1.49 0.2219 0.2307
time1*CMgirl50 1 32.1 22.99 22.99 <.0001 <.0001
CMemosup5 1 29.8 1.99 1.99 0.1580 0.1683
time1*CMemosup5 1 27.6 19.83 19.83 <.0001 0.0001
WSvictim 1 275 0.25 0.25 0.6146 0.6150
WCvictim 1 567 1.30 1.30 0.2539 0.2543
CMvictim3 1 34.6 0.20 0.20 0.6580 0.6607
CMgirl50*CMemosup5 1 37.8 8.02 8.02 0.0046 0.0074
CMemosup5*CMvictim3 1 32.2 7.27 7.27 0.0070 0.0110
girl*WCvictim 1 580 4.92 4.92 0.0266 0.0269
CMgirl50*WCvictim 1 559 0.02 0.02 0.8894 0.8894
CMemosup5*WCvictim 1 561 0.83 0.83 0.3626 0.3630
WCvictim*CMvictim3 1 552 0.32 0.32 0.5723 0.5726


Ch 11a: Add Level-3 by Within-Student Victim Cross-Level Interactions
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 18
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4604.25983954  
1 3 4077.33559692 0.00942801
2 2 4074.72224478 0.00139115
3 2 4074.08274545 0.00001525
4 1 4074.07560304 0.00000001

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.02022 0.01553 1.30 0.0965
UN(2,1) ClassID 0.01989 0.006055 3.28 0.0010
UN(2,2) ClassID 0.004390 0.004494 0.98 0.1643
UN(1,1) ClassID*StudentID 0.3957 0.04317 9.17 <.0001
UN(2,1) ClassID*StudentID -0.00779 0.02019 -0.39 0.6995
UN(2,2) ClassID*StudentID 0.04420 0.01612 2.74 0.0030
UN(3,1) ClassID*StudentID 0.02424 0.02773 0.87 0.3822
UN(3,2) ClassID*StudentID -0.01511 0.01371 -1.10 0.2704
UN(3,3) ClassID*StudentID 0.1043 0.03294 3.17 0.0008
Residual   0.2963 0.01981 14.96 <.0001

Fit Statistics
-2 Res Log Likelihood 4074.1
AIC (smaller is better) 4094.1
AICC (smaller is better) 4094.2
BIC (smaller is better) 4109.0

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 530.18 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4074.1 10 4094.1 4094.2 4099.1 4109.0 4119.0

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2107 0.05845 65.4 72.04 <.0001 0.05 4.0940 4.3274
time1 -0.1309 0.02944 78.2 -4.45 <.0001 0.05 -0.1895 -0.07235
girl 0.2519 0.06834 557 3.69 0.0002 0.05 0.1177 0.3862
time1*girl -0.00033 0.03866 541 -0.01 0.9931 0.05 -0.07628 0.07562
CMgirl50 0.5677 0.4486 32.8 1.27 0.2147 0.05 -0.3453 1.4806
time1*CMgirl50 -1.1292 0.2340 32.7 -4.82 <.0001 0.05 -1.6056 -0.6529
CMemosup5 0.1123 0.07512 30 1.49 0.1455 0.05 -0.04115 0.2657
time1*CMemosup5 0.1666 0.03822 28.5 4.36 0.0002 0.05 0.08834 0.2448
WSvictim -0.04672 0.04319 310 -1.08 0.2802 0.05 -0.1317 0.03826
WCvictim -0.05494 0.05168 572 -1.06 0.2882 0.05 -0.1564 0.04656
CMvictim3 -0.04786 0.1623 34.9 -0.29 0.7698 0.05 -0.3773 0.2816
CMgirl50*CMemosup5 -2.3416 0.8179 37.6 -2.86 0.0068 0.05 -3.9979 -0.6854
CMemosup5*CMvictim3 -0.7707 0.2824 32.1 -2.73 0.0102 0.05 -1.3458 -0.1956
girl*WCvictim -0.1528 0.07308 581 -2.09 0.0370 0.05 -0.2964 -0.00928
CMgirl50*WCvictim -0.09395 0.3709 561 -0.25 0.8001 0.05 -0.8224 0.6345
CMgirl50*WSvictim 0.1006 0.3698 262 0.27 0.7859 0.05 -0.6277 0.8288
CMemosup5*WSvictim 0.05596 0.06741 281 0.83 0.4072 0.05 -0.07673 0.1886
WSvictim*CMvictim3 0.1826 0.1464 287 1.25 0.2131 0.05 -0.1054 0.4707

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 78.2 19.79 19.79 <.0001 <.0001
girl 1 557 13.59 13.59 0.0002 0.0002
time1*girl 1 541 0.00 0.00 0.9931 0.9931
CMgirl50 1 32.8 1.60 1.60 0.2058 0.2147
time1*CMgirl50 1 32.7 23.28 23.28 <.0001 <.0001
CMemosup5 1 30 2.23 2.23 0.1351 0.1455
time1*CMemosup5 1 28.5 18.99 18.99 <.0001 0.0002
WSvictim 1 310 1.17 1.17 0.2794 0.2802
WCvictim 1 572 1.13 1.13 0.2877 0.2882
CMvictim3 1 34.9 0.09 0.09 0.7681 0.7698
CMgirl50*CMemosup5 1 37.6 8.20 8.20 0.0042 0.0068
CMemosup5*CMvictim3 1 32.1 7.45 7.45 0.0063 0.0102
girl*WCvictim 1 581 4.37 4.37 0.0365 0.0370
CMgirl50*WCvictim 1 561 0.06 0.06 0.8000 0.8001
CMgirl50*WSvictim 1 262 0.07 0.07 0.7857 0.7859
CMemosup5*WSvictim 1 281 0.69 0.69 0.4065 0.4072
WSvictim*CMvictim3 1 287 1.56 1.56 0.2121 0.2131


Eq 11a.9: Add Level-2 and Level-3 Victim by Within-Student Victim Cross-Level Interactions
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 17
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4602.64282830  
1 3 4073.07115870 0.00916304
2 2 4070.41187793 0.00124571
3 2 4069.83783700 0.00001174
4 1 4069.83238780 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.01919 0.01523 1.26 0.1039
UN(2,1) ClassID 0.02030 0.006082 3.34 0.0008
UN(2,2) ClassID 0.004708 0.004585 1.03 0.1523
UN(1,1) ClassID*StudentID 0.3954 0.04315 9.16 <.0001
UN(2,1) ClassID*StudentID -0.00748 0.02014 -0.37 0.7102
UN(2,2) ClassID*StudentID 0.04366 0.01604 2.72 0.0032
UN(3,1) ClassID*StudentID 0.01736 0.02728 0.64 0.5246
UN(3,2) ClassID*StudentID -0.01281 0.01341 -0.96 0.3393
UN(3,3) ClassID*StudentID 0.09474 0.03169 2.99 0.0014
Residual   0.2970 0.01980 15.00 <.0001

Fit Statistics
-2 Res Log Likelihood 4069.8
AIC (smaller is better) 4089.8
AICC (smaller is better) 4090.0
BIC (smaller is better) 4104.8

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 532.81 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4069.8 10 4089.8 4090.0 4094.9 4104.8 4114.8

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2116 0.05803 66.6 72.58 <.0001 0.05 4.0958 4.3274
time1 -0.1308 0.02953 76.4 -4.43 <.0001 0.05 -0.1896 -0.07199
girl 0.2517 0.06835 557 3.68 0.0003 0.05 0.1174 0.3859
time1*girl -0.00062 0.03857 542 -0.02 0.9872 0.05 -0.07639 0.07515
CMgirl50 0.5498 0.4437 32.8 1.24 0.2241 0.05 -0.3532 1.4528
time1*CMgirl50 -1.1116 0.2349 31.9 -4.73 <.0001 0.05 -1.5902 -0.6331
CMemosup5 0.1082 0.07421 29.9 1.46 0.1552 0.05 -0.04336 0.2598
time1*CMemosup5 0.1697 0.03831 27.6 4.43 0.0001 0.05 0.09119 0.2482
WSvictim -0.07518 0.04450 409 -1.69 0.0919 0.05 -0.1627 0.01230
WCvictim -0.05272 0.05167 572 -1.02 0.3081 0.05 -0.1542 0.04878
CMvictim3 -0.05383 0.1596 35.2 -0.34 0.7379 0.05 -0.3778 0.2701
CMgirl50*CMemosup5 -2.3494 0.8079 38.4 -2.91 0.0060 0.05 -3.9844 -0.7145
CMemosup5*CMvictim3 -0.7781 0.2784 32.6 -2.80 0.0086 0.05 -1.3447 -0.2115
girl*WCvictim -0.1521 0.07306 581 -2.08 0.0378 0.05 -0.2956 -0.00860
CMgirl50*WCvictim -0.09563 0.3710 561 -0.26 0.7967 0.05 -0.8243 0.6330
WSvictim*WCvictim 0.1092 0.04521 335 2.41 0.0163 0.05 0.02023 0.1981
WSvictim*CMvictim3 0.1570 0.1318 290 1.19 0.2346 0.05 -0.1024 0.4164

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 76.4 19.62 19.62 <.0001 <.0001
girl 1 557 13.56 13.56 0.0002 0.0003
time1*girl 1 542 0.00 0.00 0.9872 0.9872
CMgirl50 1 32.8 1.54 1.54 0.2153 0.2241
time1*CMgirl50 1 31.9 22.39 22.39 <.0001 <.0001
CMemosup5 1 29.9 2.13 2.13 0.1448 0.1552
time1*CMemosup5 1 27.6 19.63 19.63 <.0001 0.0001
WSvictim 1 409 2.85 2.85 0.0911 0.0919
WCvictim 1 572 1.04 1.04 0.3077 0.3081
CMvictim3 1 35.2 0.11 0.11 0.7359 0.7379
CMgirl50*CMemosup5 1 38.4 8.46 8.46 0.0036 0.0060
CMemosup5*CMvictim3 1 32.6 7.81 7.81 0.0052 0.0086
girl*WCvictim 1 581 4.33 4.33 0.0374 0.0378
CMgirl50*WCvictim 1 561 0.07 0.07 0.7966 0.7967
WSvictim*WCvictim 1 335 5.83 5.83 0.0158 0.0163
WSvictim*CMvictim3 1 290 1.42 1.42 0.2336 0.2346

Estimates
Label Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Contextual Gender Effect at Wave 1 0.5498 0.4437 32.8 1.24 0.2241 0.05 -0.3532 1.4528
Contextual Gender Effect at Wave 2 -0.5618 0.5556 30.6 -1.01 0.3198 0.05 -1.6955 0.5719
Contextual Gender Effect at Wave 3 -1.6735 0.7285 30.3 -2.30 0.0287 0.05 -3.1607 -0.1863
Between-Class Emotional Support Effect at Wave 1 0.1082 0.07421 29.9 1.46 0.1552 0.05 -0.04336 0.2598
Between-Class Emotional Support Effect at Wave 2 0.2779 0.09362 29.3 2.97 0.0059 0.05 0.08653 0.4694
Between-Class Emotional Support Effect at Wave 3 0.4477 0.1223 28.9 3.66 0.0010 0.05 0.1975 0.6979
Level-2 Victim Effect in Girls -0.2048 0.05069 575 -4.04 <.0001 0.05 -0.3044 -0.1053


PsuedoR2 (% Reduction) for CovL32x vs. CovL32xVicx

Name CovParm Subject Estimate StdErr ZValue ProbZ PseudoR2
CovL32x UN(1,1) ClassID 0.01997 0.01542 1.30 0.0976 .
CovL32x UN(2,2) ClassID 0.004571 0.004551 1.00 0.1576 .
CovL32x UN(1,1) ClassID*StudentID 0.3945 0.04308 9.16 <.0001 .
CovL32x UN(2,2) ClassID*StudentID 0.04425 0.01610 2.75 0.0030 .
CovL32x UN(3,3) ClassID*StudentID 0.1023 0.03241 3.16 0.0008 .
CovL32x Residual   0.2963 0.01977 14.99 <.0001 .
CovL32xVicx UN(1,1) ClassID 0.01919 0.01523 1.26 0.1039 0.039261
CovL32xVicx UN(2,2) ClassID 0.004708 0.004585 1.03 0.1523 -0.029910
CovL32xVicx UN(1,1) ClassID*StudentID 0.3954 0.04315 9.16 <.0001 -0.002315
CovL32xVicx UN(2,2) ClassID*StudentID 0.04366 0.01604 2.72 0.0032 0.013294
CovL32xVicx UN(3,3) ClassID*StudentID 0.09474 0.03169 2.99 0.0014 0.073928
CovL32xVicx Residual   0.2970 0.01980 15.00 <.0001 -0.002208


Total R2 (% Reduction) for PredEmpty vs. FinalC

Name PredCorr TotalR2 TotalR2Diff
PredEmpty . . .
FinalC 0.32317 0.10444 .


Ch 11a: Add Level-2 and Level-3 Gender by Within-Student Victim Cross-Level Interactions
Predicting Student Closeness

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method REML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 19
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Res Log Like Criterion
0 1 4604.63764547  
1 3 4075.73256653 0.00901264
2 2 4073.06676849 0.00112334
3 2 4072.54940513 0.00000986
4 1 4072.54479831 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.01918 0.01523 1.26 0.1040
UN(2,1) ClassID 0.02032 0.006078 3.34 0.0008
UN(2,2) ClassID 0.004686 0.004590 1.02 0.1536
UN(1,1) ClassID*StudentID 0.3961 0.04318 9.17 <.0001
UN(2,1) ClassID*StudentID -0.00792 0.02016 -0.39 0.6945
UN(2,2) ClassID*StudentID 0.04378 0.01605 2.73 0.0032
UN(3,1) ClassID*StudentID 0.01910 0.02738 0.70 0.4854
UN(3,2) ClassID*StudentID -0.01336 0.01348 -0.99 0.3219
UN(3,3) ClassID*StudentID 0.09598 0.03194 3.00 0.0013
Residual   0.2970 0.01979 15.01 <.0001

Fit Statistics
-2 Res Log Likelihood 4072.5
AIC (smaller is better) 4092.5
AICC (smaller is better) 4092.7
BIC (smaller is better) 4107.5

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 532.09 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4072.5 10 4092.5 4092.7 4097.6 4107.5 4117.5

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2137 0.05812 66.9 72.50 <.0001 0.05 4.0976 4.3297
time1 -0.1331 0.02965 77.3 -4.49 <.0001 0.05 -0.1921 -0.07402
girl 0.2466 0.06862 555 3.59 0.0004 0.05 0.1119 0.3814
time1*girl 0.004058 0.03893 533 0.10 0.9170 0.05 -0.07242 0.08054
CMgirl50 0.5675 0.4447 33.1 1.28 0.2108 0.05 -0.3372 1.4721
time1*CMgirl50 -1.1249 0.2356 32.3 -4.77 <.0001 0.05 -1.6047 -0.6452
CMemosup5 0.1090 0.07424 29.9 1.47 0.1525 0.05 -0.04263 0.2606
time1*CMemosup5 0.1691 0.03831 27.5 4.42 0.0001 0.05 0.09059 0.2477
WSvictim -0.04630 0.05765 307 -0.80 0.4226 0.05 -0.1597 0.06715
WCvictim -0.05260 0.05168 572 -1.02 0.3092 0.05 -0.1541 0.04890
CMvictim3 -0.05000 0.1598 35.3 -0.31 0.7562 0.05 -0.3744 0.2744
CMgirl50*CMemosup5 -2.3737 0.8084 38.5 -2.94 0.0056 0.05 -4.0096 -0.7379
CMemosup5*CMvictim3 -0.7805 0.2784 32.7 -2.80 0.0085 0.05 -1.3472 -0.2137
girl*WCvictim -0.1524 0.07307 582 -2.09 0.0375 0.05 -0.2959 -0.00888
CMgirl50*WCvictim -0.09677 0.3710 561 -0.26 0.7943 0.05 -0.8254 0.6319
WSvictim*WCvictim 0.1090 0.04528 333 2.41 0.0166 0.05 0.01994 0.1981
WSvictim*CMvictim3 0.1648 0.1364 275 1.21 0.2279 0.05 -0.1037 0.4332
girl*WSvictim -0.06335 0.07267 265 -0.87 0.3842 0.05 -0.2064 0.07974
CMgirl50*WSvictim 0.1865 0.3665 250 0.51 0.6113 0.05 -0.5352 0.9082

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 77.3 20.14 20.14 <.0001 <.0001
girl 1 555 12.92 12.92 0.0003 0.0004
time1*girl 1 533 0.01 0.01 0.9170 0.9170
CMgirl50 1 33.1 1.63 1.63 0.2019 0.2108
time1*CMgirl50 1 32.3 22.80 22.80 <.0001 <.0001
CMemosup5 1 29.9 2.16 2.16 0.1420 0.1525
time1*CMemosup5 1 27.5 19.49 19.49 <.0001 0.0001
WSvictim 1 307 0.64 0.64 0.4219 0.4226
WCvictim 1 572 1.04 1.04 0.3088 0.3092
CMvictim3 1 35.3 0.10 0.10 0.7544 0.7562
CMgirl50*CMemosup5 1 38.5 8.62 8.62 0.0033 0.0056
CMemosup5*CMvictim3 1 32.7 7.86 7.86 0.0051 0.0085
girl*WCvictim 1 582 4.35 4.35 0.0370 0.0375
CMgirl50*WCvictim 1 561 0.07 0.07 0.7942 0.7943
WSvictim*WCvictim 1 333 5.80 5.80 0.0161 0.0166
WSvictim*CMvictim3 1 275 1.46 1.46 0.2269 0.2279
girl*WSvictim 1 265 0.76 0.76 0.3834 0.3842
CMgirl50*WSvictim 1 250 0.26 0.26 0.6108 0.6113


Eq 11a.9: Add Level-2 and Level-3 Victim by Within-Student Victim Cross-Level Interactions
Predicting Student Closeness using ML instead of REML

The Mixed Procedure

Model Information
Data Set WORK.CHAPTER11A
Dependent Variable close
Covariance Structure Unstructured
Subject Effects ClassID, ClassID*StudentID
Estimation Method ML
Residual Variance Method Profile
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Satterthwaite

Dimensions
Covariance Parameters 10
Columns in X 17
Columns in Z Per Subject 77
Subjects 33
Max Obs Per Subject 73

Number of Observations
Number of Observations Read 1731
Number of Observations Used 1731
Number of Observations Not Used 0

Iteration History
Iteration Evaluations -2 Log Like Criterion
0 1 4546.94779663  
1 3 4023.14229527 0.01309032
2 2 4020.52227626 0.00261762
3 2 4019.48317158 0.00006959
4 1 4019.45316247 0.00000016
5 1 4019.45309404 0.00000000

Convergence criteria met.

Covariance Parameter Estimates
Cov Parm Subject Estimate Standard Error Z Value Pr Z
UN(1,1) ClassID 0.01081 0.01182 0.91 0.1803
UN(2,1) ClassID 0.01854 0.005056 3.67 0.0002
UN(2,2) ClassID 0.003373 0.003999 0.84 0.1995
UN(1,1) ClassID*StudentID 0.3896 0.04263 9.14 <.0001
UN(2,1) ClassID*StudentID -0.00630 0.01996 -0.32 0.7523
UN(2,2) ClassID*StudentID 0.04293 0.01594 2.69 0.0035
UN(3,1) ClassID*StudentID 0.01683 0.02684 0.63 0.5307
UN(3,2) ClassID*StudentID -0.01305 0.01321 -0.99 0.3232
UN(3,3) ClassID*StudentID 0.09076 0.03109 2.92 0.0018
Residual   0.2971 0.01978 15.02 <.0001

Fit Statistics
-2 Log Likelihood 4019.5
AIC (smaller is better) 4073.5
AICC (smaller is better) 4074.3
BIC (smaller is better) 4113.9

Null Model Likelihood Ratio Test
DF Chi-Square Pr > ChiSq
9 527.49 <.0001

Information Criteria
Neg2LogLike Parms AIC AICC HQIC BIC CAIC
4019.5 27 4073.5 4074.3 4087.0 4113.9 4140.9

Solution for Fixed Effects
Effect Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Intercept 4.2115 0.05492 88.1 76.69 <.0001 0.05 4.1024 4.3206
time1 -0.1309 0.02875 89.5 -4.55 <.0001 0.05 -0.1880 -0.07378
girl 0.2518 0.06802 562 3.70 0.0002 0.05 0.1181 0.3854
time1*girl -0.00085 0.03846 545 -0.02 0.9824 0.05 -0.07639 0.07469
CMgirl50 0.5372 0.4103 39.9 1.31 0.1979 0.05 -0.2921 1.3665
time1*CMgirl50 -1.1124 0.2254 35.4 -4.94 <.0001 0.05 -1.5697 -0.6550
CMemosup5 0.1089 0.06838 36 1.59 0.1201 0.05 -0.02981 0.2475
time1*CMemosup5 0.1701 0.03667 30.5 4.64 <.0001 0.05 0.09530 0.2450
WSvictim -0.07450 0.04420 409 -1.69 0.0927 0.05 -0.1614 0.01239
WCvictim -0.05196 0.05138 579 -1.01 0.3123 0.05 -0.1529 0.04895
CMvictim3 -0.05810 0.1480 43.8 -0.39 0.6964 0.05 -0.3563 0.2401
CMgirl50*CMemosup5 -2.3592 0.7511 47.9 -3.14 0.0029 0.05 -3.8693 -0.8491
CMemosup5*CMvictim3 -0.7757 0.2574 40.5 -3.01 0.0044 0.05 -1.2957 -0.2557
girl*WCvictim -0.1534 0.07254 591 -2.11 0.0349 0.05 -0.2959 -0.01093
CMgirl50*WCvictim -0.09283 0.3694 566 -0.25 0.8017 0.05 -0.8184 0.6327
WSvictim*WCvictim 0.1102 0.04487 334 2.46 0.0146 0.05 0.02194 0.1984
WSvictim*CMvictim3 0.1569 0.1308 289 1.20 0.2312 0.05 -0.1005 0.4143

Type 3 Tests of Fixed Effects
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F
time1 1 89.5 20.73 20.73 <.0001 <.0001
girl 1 562 13.70 13.70 0.0002 0.0002
time1*girl 1 545 0.00 0.00 0.9824 0.9824
CMgirl50 1 39.9 1.71 1.71 0.1904 0.1979
time1*CMgirl50 1 35.4 24.36 24.36 <.0001 <.0001
CMemosup5 1 36 2.53 2.53 0.1114 0.1201
time1*CMemosup5 1 30.5 21.53 21.53 <.0001 <.0001
WSvictim 1 409 2.84 2.84 0.0919 0.0927
WCvictim 1 579 1.02 1.02 0.3118 0.3123
CMvictim3 1 43.8 0.15 0.15 0.6945 0.6964
CMgirl50*CMemosup5 1 47.9 9.87 9.87 0.0017 0.0029
CMemosup5*CMvictim3 1 40.5 9.08 9.08 0.0026 0.0044
girl*WCvictim 1 591 4.47 4.47 0.0345 0.0349
CMgirl50*WCvictim 1 566 0.06 0.06 0.8016 0.8017
WSvictim*WCvictim 1 334 6.03 6.03 0.0140 0.0146
WSvictim*CMvictim3 1 289 1.44 1.44 0.2303 0.2312