187
SEM Simulations SIMULATION #1 Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior) Data Set: Model 3_Intention (Att, PB, and Behavior in relation to Intention) SAS SYNTAX: proc calis data =SASUSER.SEM_INTDATA; path attitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, pastbeh ---> X10 X11 X12 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention beh, pastbeh---> intention beh, subnorms ---> intention, intention ---> beh; run ; SAS LOG: 1 proc calis data=SASUSER.SEM_INTDATA; 2 path 3 attitude ---> X1 X2 X3 = lam1 lam2 lam3, 4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3, 5 intention ---> X7 X8 X9 = lam1 lam2 lam3, 6 pastbeh ---> X10 X11 X12 = lam1 lam2 lam3, 7 beh ---> X13 X14 X15 = lam1 lam2 lam3, 8 attitude ---> intention beh, 9 pastbeh---> intention beh, 10 subnorms ---> intention, 11 intention ---> beh; 12 run; NOTE: Convergence criterion (GCONV=1E-8) satisfied. NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates. WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse. NOTE: PROCEDURE CALIS used (Total process time): real time 1.20 seconds cpu time 0.87 seconds SAS OUTPUT:

tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

SEM Simulations

SIMULATION #1

Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Data Set: Model 3_Intention (Att, PB, and Behavior in relation to Intention)

SAS SYNTAX:

proc calis data=SASUSER.SEM_INTDATA;path attitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, pastbeh ---> X10 X11 X12 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention beh, pastbeh---> intention beh, subnorms ---> intention, intention ---> beh; run;

SAS LOG:

1 proc calis data=SASUSER.SEM_INTDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 pastbeh ---> X10 X11 X12 = lam1 lam2 lam3,7 beh ---> X13 X14 X15 = lam1 lam2 lam3,8 attitude ---> intention beh,9 pastbeh---> intention beh,10 subnorms ---> intention,11 intention ---> beh;12 run;

NOTE: Convergence criterion (GCONV=1E-8) satisfied.NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.NOTE: PROCEDURE CALIS used (Total process time): real time 1.20 seconds cpu time 0.87 seconds

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_INTDATA

Page 2: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Modeling Information

N Records Read

228

N Records Used 228

N Obs 228

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X10 X11 X12 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Latent beh intention

Exogenous Manifest

Latent attitude pastbeh subnorms

Number of Endogenous Variables = 17Number of Exogenous Variables = 3

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 -0.18557 2.22965

X2 X2 -2.20722 2.19830

X3 X3 -3.79376 1.81540

X4 X4 -0.81200 2.95223

X5 X5 0.08676 2.43522

X6 X6 2.35058 2.15656

X7 X7 -0.48882 1.30821

X8 X8 -0.97984 1.33089

X9 X9 1.69498 1.31175

X10

X10 0.34446 4.10414

Page 3: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simple Statistics

Variable Mean Std Dev

X11

X11 -7.14348 3.74671

X12

X12 3.66854 3.40156

X13

X13 0.58236 1.37847

X14

X14 0.46012 1.44859

X15

X15 0.54218 1.28555

Page 4: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 -0.59768

2 lam2 0.88609 0.15125

3 lam3 0.76996 0.01355

4 _Parm1 0.13192 -0.09498

5 _Parm2 -0.03676 -0.01859

6 _Parm3 0.08253 -0.16175

7 _Parm4 0.02599 0.02477

8 _Parm5 0.14076 0.09933

9 _Parm6 0.23829 0.00267

10 _Add01 4.24777 -0.06989

11 _Add02 6.86760 -0.18432

12 _Add03 16.65435 -0.07484

13 _Add04 0.72357 -0.21081

14 _Add05 1.49735 0.12793

15 _Add06 0.77741 -0.25448

16 _Add07 1.84807 0.12764

17 _Add08 0.53817 -0.34487

18 _Add09 0.57932 -0.54889

19 _Add10 0.76450 -0.07901

Page 5: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add11 1.02778 0.00209

21 _Add12 1.15930 0.01920

22 _Add13 0.18964 -0.97008

23 _Add14 0.96159 -0.50598

24 _Add15 1.69717 0.13771

25 _Add16 0.88437 -0.02383

26 _Add17 1.30086 0.00451

27 _Add18 1.05044 0.00681

28 _Add19 0.06455 -0.05439

29 _Add20 0.96392 0.01076

30 _Add21 4.65652 0.07849

31 _Add22 7.21990 0.02319

32 _Add23 9.84877 0.20740

Value of Objective Function = 0.9781959536

Page 6: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 32

Functions (Observations) 120

Optimization Start

Active Constraints 0 Objective Function 0.9781959536

Max Abs Gradient Element 0.9700773661 Radius 4.9416960031

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 * 0 4 0 0.46272 0.5155 0.3040 111E-16

0.751

2 * 0 6 0 0.45281 0.00992 0.0453 111E-16

0.839

3 * 0 8 0 0.45188 0.000925

0.0184 111E-16

0.938

4 * 0 10 0 0.45182 0.000065

0.00225 111E-16

0.998

5 * 0 12 0 0.45181 5.546E-6

0.00130 111E-16

1.058

6 * 0 14 0 0.45181 5.279E-7

0.000196

111E-16

1.112

7 * 0 16 0 0.45181 5.202E- 0.00011 111E- 1.160

Page 7: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

8 4 16

8 * 0 18 0 0.45181 5.25E-9 0.000019

111E-16

1.200

9 * 0 20 0 0.45181 5.38E-10

0.000011

111E-16

1.232

Optimization Results

Iterations 9 Function Calls 23

Jacobian Calls 11 Active Constraints 0

Objective Function 0.4518098078 Max Abs Gradient Element 0.0000111963

Lambda 1.110223E-14 Actual Over Pred Change 1.2323157247

Radius 5692.8337955

Convergence criterion (GCONV=1E-8) satisfied.

Note: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.

Warning: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.

Note: Covariance matrix for the estimates is not full rank.

Note: The variance of some parameter estimates is zero or some parameter estimates are linearly related to other parameter estimates as shown in the following equations:

_Ad

=141.75

+ 0.14

*lam

+0.12

*lam

+0.11

*lam

- 1.24

*_Ad

- 2.00

*_Ad

- 4.85

*_Ad

- 0.01

*_Ad

- 0.27

*_Ad

- 1.34

*_Ad

- 2.85

*_Ad

Page 8: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

d22

7969

3704

1 8427

2 1948

3 0364

d01

4088

d02

3819

d03

8282

d19

4115

d20

7196

d21

5816

d23

Page 9: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 15

N Moments 120

N Parameters 32

N Active Constraints 0

Baseline Model Function Value 14.1366

Baseline Model Chi-Square 3209.0143

Baseline Model Chi-Square DF 105

Pr > Baseline Model Chi-Square <.0001

Absolute Index Fit Function 0.4518

Chi-Square 102.5608

Chi-Square DF 88

Pr > Chi-Square 0.1374

Z-Test of Wilson & Hilferty 1.0923

Hoelter Critical N 246

Root Mean Square Residual (RMSR) 0.2457

Standardized RMSR (SRMSR) 0.0449

Goodness of Fit Index (GFI) 0.9430

Parsimony Index Adjusted GFI (AGFI) 0.9223

Parsimonious GFI 0.7903

RMSEA Estimate 0.0270

RMSEA Lower 90% Confidence Limit 0.0000

RMSEA Upper 90% Confidence Limit 0.0469

Probability of Close Fit 0.9744

ECVI Estimate 0.7551

Page 10: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.7204

ECVI Upper 90% Confidence Limit 0.8869

Akaike Information Criterion 166.5608

Bozdogan CAIC 308.2999

Schwarz Bayesian Criterion 276.2999

McDonald Centrality 0.9686

Incremental Index

Bentler Comparative Fit Index 0.9953

Bentler-Bonett NFI 0.9680

Bentler-Bonett Non-normed Index 0.9944

Bollen Normed Index Rho1 0.9619

Bollen Non-normed Index Delta2 0.9953

James et al. Parsimonious NFI 0.8113

Page 11: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter Estimate

StandardError

t Value

attitude --->

X1 lam1 0.98962 0.04729 20.92755

attitude --->

X2 lam2 0.88442 0.04285 20.64206

attitude --->

X3 lam3 0.77094 0.03780 20.39521

subnorms --->

X4 lam1 0.98962 0.04729 20.92755

subnorms --->

X5 lam2 0.88442 0.04285 20.64206

subnorms --->

X6 lam3 0.77094 0.03780 20.39521

intention --->

X7 lam1 0.98962 0.04729 20.92755

intention --->

X8 lam2 0.88442 0.04285 20.64206

intention --->

X9 lam3 0.77094 0.03780 20.39521

pastbeh --->

X10 lam1 0.98962 0.04729 20.92755

pastbeh --->

X11 lam2 0.88442 0.04285 20.64206

pastbeh --->

X12 lam3 0.77094 0.03780 20.39521

beh --->

X13 lam1 0.98962 0.04729 20.92755

beh --->

X14 lam2 0.88442 0.04285 20.64206

beh --- X15 lam3 0.77094 0.03780 20.39521

Page 12: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

PATH List

Path Parameter Estimate

StandardError

t Value

>

attitude --->

intention _Parm1 0.14782 0.05658 2.61253

attitude --->

beh _Parm2 0.01303 0.14928 0.08729

pastbeh --->

intention _Parm3 0.09604 0.03592 2.67378

pastbeh --->

beh _Parm4 0.06159 0.09812 0.62770

subnorms --->

intention _Parm5 0.10752 0.05705 1.88448

intention --->

beh _Parm6 -0.01311 0.62240 -0.02107

Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 4.27091 0.30570 13.97076

subnorms _Add02 6.90063 0.34813 19.82203

pastbeh _Add03 16.71304 0.31915 52.36759

Error X1 _Add04 0.97754 0.13465 7.26007

X2 _Add05 1.04880 0.12871 8.14852

X3 _Add06 0.98239 0.11372 8.63890

X4 _Add07 1.17038 0.14733 7.94382

X5 _Add08 0.74346 0.10288 7.22630

X6 _Add09 0.89916 0.10541 8.53023

X7 _Add10 0.80976 0.09318 8.69018

X8 _Add11 1.03032 0.10678 9.64928

X9 _Add12 1.12275 0.11174 10.04817

X10 _Add13 0.80383 0.14013 5.73653

Page 13: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

X11 _Add14 1.14986 0.14682 7.83160

X12 _Add15 1.04377 0.12527 8.33188

X13 _Add16 0.94555 0.13439 7.03580

X14 _Add17 1.27833 0.14932 8.56084

X15 _Add18 1.03177 0.11878 8.68661

intention _Add19 0.06295 0.05408 1.16402

beh _Add20 0.94385 0.16141 5.84769

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add21 4.63877 0.22893 20.26319

pastbeh attitude _Add22 7.25955 0.27290 26.60123

pastbeh subnorms _Add23 9.83336 0.21634 45.45428

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 0.97754 5.16028 0.8106

X10 0.80383 17.17183 0.9532

X11 1.14986 14.22276 0.9192

X12 1.04377 10.97709 0.9049

X13 0.94555 1.93776 0.5120

X14 1.27833 2.07080 0.3827

X15 1.03177 1.63392 0.3685

X2 1.04880 4.38949 0.7611

X3 0.98239 3.52079 0.7210

X4 1.17038 7.92854 0.8524

X5 0.74346 6.14112 0.8789

X6 0.89916 5.00051 0.8202

Page 14: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X7 0.80976 1.73705 0.5338

X8 1.03032 1.77093 0.4182

X9 1.12275 1.68550 0.3339

beh 0.94385 1.01313 0.0684

intention 0.06295 0.94683 0.9335

Page 15: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter Estimate StandardError

t Value

attitude ---> X1 lam1 0.90031 0.01525 59.03827

attitude ---> X2 lam2 0.87239 0.01731 50.40649

attitude ---> X3 lam3 0.84910 0.01906 44.56050

subnorms

---> X4 lam1 0.92325 0.01123 82.23967

subnorms

---> X5 lam2 0.93752 0.00996 94.09365

subnorms

---> X6 lam3 0.90564 0.01293 70.06525

intention ---> X7 lam1 0.73064 0.03338 21.88720

intention ---> X8 lam2 0.64669 0.03455 18.71960

intention ---> X9 lam3 0.57782 0.03443 16.78082

pastbeh ---> X10 lam1 0.97631 0.00465 209.80735

pastbeh ---> X11 lam2 0.95872 0.00637 150.59335

pastbeh ---> X12 lam3 0.95127 0.00718 132.48656

beh ---> X13 lam1 0.71557 0.03834 18.66307

beh ---> X14 lam2 0.61862 0.03661 16.89663

beh ---> X15 lam3 0.60707 0.03651 16.62549

attitude ---> intention

_Parm1 0.31395 0.11933 2.63092

attitude ---> beh _Parm2 0.02675 0.30648 0.08729

pastbeh ---> intention

_Parm3 0.40349 0.15011 2.68791

pastbeh ---> beh _Parm4 0.25014 0.39807 0.62839

subnorms

---> intention

_Parm5 0.29025 0.15354 1.89036

Page 16: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for PATH List

Path Parameter Estimate StandardError

t Value

intention ---> beh _Parm6 -0.01268 0.60169 -0.02107

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 1.00000

subnorms _Add02 1.00000

pastbeh _Add03 1.00000

Error X1 _Add04 0.18944 0.02746 6.89884

X2 _Add05 0.23893 0.03020 7.91243

X3 _Add06 0.27903 0.03236 8.62275

X4 _Add07 0.14762 0.02073 7.12113

X5 _Add08 0.12106 0.01868 6.48015

X6 _Add09 0.17981 0.02341 7.68034

X7 _Add10 0.46617 0.04878 9.55657

X8 _Add11 0.58179 0.04468 13.02105

X9 _Add12 0.66612 0.03979 16.73996

X10 _Add13 0.04681 0.00909 5.15184

X11 _Add14 0.08085 0.01221 6.62292

X12 _Add15 0.09509 0.01366 6.96070

X13 _Add16 0.48796 0.05487 8.89266

X14 _Add17 0.61731 0.04530 13.62804

X15 _Add18 0.63147 0.04433 14.24374

intention _Add19 0.06648 0.05385 1.23461

beh _Add20 0.93162 0.03997 23.30951

Page 17: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add21 0.85447 0.02376 35.95744

pastbeh attitude _Add22 0.85925 0.02195 39.13942

pastbeh subnorms _Add23 0.91565 0.01378 66.42591

Simulation #2

Test: Model 2 (Intention Partially Mediates Attitude Effects on Behavior)

Data Set: Model 3_Intention (Att, PB, and Behavior in relation to Intention)

SAS SYNTAX:

proc calis data=SASUSER.SEM_INTDATA; path attitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention beh, subnorms ---> intention, intention ---> beh;run;

SAS LOG:

1 proc calis data=SASUSER.SEM_INTDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 beh ---> X13 X14 X15 = lam1 lam2 lam3,7 attitude ---> intention beh,8 subnorms ---> intention,9 intention ---> beh;10 run;

NOTE: Convergence criterion (GCONV=1E-8) satisfied.NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.NOTE: PROCEDURE CALIS used (Total process time): real time 1.10 seconds cpu time 0.67 seconds

Page 18: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_INTDATA

N Records Read

228

N Records Used 228

N Obs 228

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Latent beh intention

Exogenous Manifest

Latent attitude subnorms

Number of Endogenous Variables = 14Number of Exogenous Variables = 2

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 -0.18557 2.22965

X2 X2 -2.20722 2.19830

X3 X3 -3.79376 1.81540

X4 X4 -0.81200 2.95223

X5 X5 0.08676 2.43522

X6 X6 2.35058 2.15656

Page 19: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simple Statistics

Variable Mean Std Dev

X7 X7 -0.48882 1.30821

X8 X8 -0.97984 1.33089

X9 X9 1.69498 1.31175

X13

X13 0.58236 1.37847

X14

X14 0.46012 1.44859

X15

X15 0.54218 1.28555

Page 20: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 0.00470

2 lam2 0.87495 -0.17665

3 lam3 0.72456 0.00476

4 _Parm1 0.17144 0.03046

5 _Parm2 -0.04647 -0.0004326

6 _Parm3 0.23105 0.20110

7 _Parm4 0.36586 0.00535

8 _Add01 4.43522 -0.07750

9 _Add02 7.16990 -0.05049

10 _Add03 0.53613 -0.35375

11 _Add04 1.43723 0.10279

12 _Add05 0.96724 -0.09507

13 _Add06 1.54578 0.07602

14 _Add07 0.44156 -0.33061

15 _Add08 0.88661 -0.05908

16 _Add09 0.72424 -0.13932

17 _Add10 1.01554 -0.01629

18 _Add11 1.20242 0.02992

19 _Add12 0.84194 -0.02563

Page 21: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add13 1.28831 -0.00516

21 _Add14 1.09709 0.01834

22 _Add15 0.09057 -0.13502

23 _Add16 1.01705 0.01858

24 _Add17 4.84064 0.12929

Value of Objective Function = 0.4751269366

Page 22: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 24

Functions (Observations) 78

Optimization Start

Active Constraints 0 Objective Function 0.4751269366

Max Abs Gradient Element 0.353754238

Radius 1.4982740861

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 * 0 4 0 0.32317 0.1520 0.1762 111E-16

0.872

2 * 0 6 0 0.31992 0.00325 0.00785 111E-16

0.976

3 * 0 8 0 0.31975 0.000173

0.00823 111E-16

1.088

4 * 0 10 0 0.31973 0.000016

0.00106 111E-16

1.234

5 * 0 12 0 0.31973 1.857E-6

0.000683

111E-16

1.318

6 * 0 14 0 0.31973 2.387E-7

0.000185

111E-16

1.350

Page 23: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

7 * 0 16 0 0.31973 3.155E-8

0.000079

111E-16

1.361

8 * 0 18 0 0.31973 4.206E-9

0.000026

111E-16

1.364

9 * 0 20 0 0.31973 5.62E-10

0.000010

111E-16

1.365

Optimization Results

Iterations 9 Function Calls 23

Jacobian Calls 11 Active Constraints 0

Objective Function 0.3197282375 Max Abs Gradient Element 0.0000100744

Lambda 1.110223E-14 Actual Over Pred Change 1.3652723962

Radius 1726.0117472

Convergence criterion (GCONV=1E-8) satisfied.

Note: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.

Warning: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.

Note: Covariance matrix for the estimates is not full rank.

Note: The variance of some parameter estimates is zero or some parameter estimates are linearly related to other parameter estimates as shown in the following equations:

_Add

= 27.406

+ 0.1587

* lam

+ 0.1398

* lam

+ 0.1160

* lam

- 1.4112

* _Add

- 2.2908

* _Add

- 0.0247

* _Add

- 0.3111

* _Add

Page 24: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

17 744 55 1 64 2 37 3 96 01 38 02 81 15 63 16

Page 25: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 12

N Moments 78

N Parameters 24

N Active Constraints 0

Baseline Model Function Value 8.0815

Baseline Model Chi-Square 1834.4924

Baseline Model Chi-Square DF 66

Pr > Baseline Model Chi-Square <.0001

Absolute Index Fit Function 0.3197

Chi-Square 72.5783

Chi-Square DF 54

Pr > Chi-Square 0.0466

Z-Test of Wilson & Hilferty 1.6788

Hoelter Critical N 226

Root Mean Square Residual (RMSR) 0.1760

Standardized RMSR (SRMSR) 0.0458

Goodness of Fit Index (GFI) 0.9492

Parsimony Index Adjusted GFI (AGFI) 0.9267

Parsimonious GFI 0.7767

RMSEA Estimate 0.0389

RMSEA Lower 90% Confidence Limit 0.0051

RMSEA Upper 90% Confidence Limit 0.0605

Probability of Close Fit 0.7811

ECVI Estimate 0.5440

Page 26: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.4766

ECVI Upper 90% Confidence Limit 0.6623

Akaike Information Criterion 120.5783

Bozdogan CAIC 226.8826

Schwarz Bayesian Criterion 202.8826

McDonald Centrality 0.9601

Incremental Index

Bentler Comparative Fit Index 0.9895

Bentler-Bonett NFI 0.9604

Bentler-Bonett Non-normed Index 0.9872

Bollen Normed Index Rho1 0.9516

Bollen Non-normed Index Delta2 0.9896

James et al. Parsimonious NFI 0.7858

Page 27: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter Estimate

StandardError

t Value

attitude --->

X1 lam1 0.99355 0.04917 20.20479

attitude --->

X2 lam2 0.87532 0.04333 20.19988

attitude --->

X3 lam3 0.72620 0.03782 19.20317

subnorms --->

X4 lam1 0.99355 0.04917 20.20479

subnorms --->

X5 lam2 0.87532 0.04333 20.19988

subnorms --->

X6 lam3 0.72620 0.03782 19.20317

intention --->

X7 lam1 0.99355 0.04917 20.20479

intention --->

X8 lam2 0.87532 0.04333 20.19988

intention --->

X9 lam3 0.72620 0.03782 19.20317

beh --->

X13 lam1 0.99355 0.04917 20.20479

beh --->

X14 lam2 0.87532 0.04333 20.19988

beh --->

X15 lam3 0.72620 0.03782 19.20317

attitude --->

intention _Parm1 0.19564 0.05475 3.57333

attitude --->

beh _Parm2 -0.02080 0.14213 -0.14635

subnorms --- intention _Parm3 0.21116 0.04231 4.99117

Page 28: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

PATH List

Path Parameter Estimate

StandardError

t Value

>

intention --->

beh _Parm4 0.31369 0.30746 1.02028

Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 4.41621 0.26026 16.96821

subnorms _Add02 7.16846 0.21257 33.72261

Error X1 _Add03 0.96584 0.14023 6.88766

X2 _Add04 1.03729 0.13057 7.94439

X3 _Add05 0.99699 0.11407 8.74016

X4 _Add06 1.11778 0.15457 7.23174

X5 _Add07 0.72936 0.11042 6.60534

X6 _Add08 0.93293 0.10954 8.51679

X7 _Add09 0.80574 0.09482 8.49746

X8 _Add10 1.03587 0.10815 9.57782

X9 _Add11 1.14412 0.11358 10.07351

X13 _Add12 0.90715 0.13566 6.68691

X14 _Add13 1.28975 0.15108 8.53703

X15 _Add14 1.06097 0.11933 8.89129

intention _Add15 0.07754 0.05699 1.36074

beh _Add16 0.97369 0.16645 5.84987

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate

StandardError

t Value

subnorms

attitude

_Add17 4.81191 0.12491 38.52294

Page 29: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 0.96584 5.32526 0.8186

X13 0.90715 1.93960 0.5323

X14 1.28975 2.09110 0.3832

X15 1.06097 1.61255 0.3421

X2 1.03729 4.42092 0.7654

X3 0.99699 3.32597 0.7002

X4 1.11778 8.19406 0.8636

X5 0.72936 6.22172 0.8828

X6 0.93293 4.71336 0.8021

X7 0.80574 1.75713 0.5414

X8 1.03587 1.77430 0.4162

X9 1.14412 1.65239 0.3076

beh 0.97369 1.04590 0.0690

intention 0.07754 0.96379 0.9195

Page 30: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter Estimate

StandardError

t Value

attitude --->

X1 lam1 0.90478 0.01550 58.39169

attitude --->

X2 lam2 0.87485 0.01770 49.43387

attitude --->

X3 lam3 0.83680 0.02076 40.30198

subnorms --->

X4 lam1 0.92929 0.01138 81.63531

subnorms --->

X5 lam2 0.93956 0.01050 89.44171

subnorms --->

X6 lam3 0.89558 0.01464 61.16794

intention --->

X7 lam1 0.73583 0.03368 21.84465

intention --->

X8 lam2 0.64512 0.03508 18.39125

intention --->

X9 lam3 0.55462 0.03500 15.84632

beh --->

X13 lam1 0.72959 0.03929 18.56897

beh --->

X14 lam2 0.61905 0.03715 16.66337

beh --->

X15 lam3 0.58485 0.03697 15.82024

attitude --->

intention _Parm1 0.41878 0.11569 3.61972

attitude --->

beh _Parm2 -0.04274 0.29204 -0.14636

subnorms --- intention _Parm3 0.57589 0.11256 5.11645

Page 31: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for PATH List

Path Parameter Estimate

StandardError

t Value

>

intention --->

beh _Parm4 0.30113 0.29334 1.02656

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 1.00000

subnorms _Add02 1.00000

Error X1 _Add03 0.18137 0.02804 6.46839

X2 _Add04 0.23463 0.03097 7.57727

X3 _Add05 0.29976 0.03475 8.62624

X4 _Add06 0.13641 0.02116 6.44760

X5 _Add07 0.11723 0.01974 5.93874

X6 _Add08 0.19793 0.02623 7.54744

X7 _Add09 0.45855 0.04957 9.25019

X8 _Add10 0.58382 0.04526 12.89944

X9 _Add11 0.69240 0.03882 17.83501

X13 _Add12 0.46770 0.05733 8.15764

X14 _Add13 0.61678 0.04600 13.40950

X15 _Add14 0.65795 0.04324 15.21527

intention _Add15 0.08046 0.05531 1.45455

beh _Add16 0.93095 0.04355 21.37538

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add17 0.85522 0.02364 36.18429

Page 32: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

SIMULATION #3

Tested: Model 1 (Theory of Reasoned Action)

Data Set: Model 3_Intention (Att, PB, and Behavior in relation to Intention)

SAS SYNTAX:

proc calis data=SASUSER.SEM_INTDATA;pathattitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention, subnorms ---> intention, intention ---> beh;run;

SAS LOG:

1 proc calis data=SASUSER.SEM_INTDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 beh ---> X13 X14 X15 = lam1 lam2 lam3,7 attitude ---> intention,8 subnorms ---> intention,9 intention ---> beh;10 run;

NOTE: Convergence criterion (GCONV=1E-8) satisfied.NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.NOTE: PROCEDURE CALIS used (Total process time): real time 1.09 seconds cpu time 0.74 seconds

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_INTDATA

N Records Read

228

N Records Used 228

N Obs 228

Page 33: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Modeling Information

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Latent beh intention

Exogenous Manifest

Latent attitude subnorms

Number of Endogenous Variables = 14Number of Exogenous Variables = 2

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 -0.18557 2.22965

X2 X2 -2.20722 2.19830

X3 X3 -3.79376 1.81540

X4 X4 -0.81200 2.95223

X5 X5 0.08676 2.43522

X6 X6 2.35058 2.15656

X7 X7 -0.48882 1.30821

X8 X8 -0.97984 1.33089

X9 X9 1.69498 1.31175

X13

X13 0.58236 1.37847

X14

X14 0.46012 1.44859

X15

X15 0.54218 1.28555

Page 34: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)
Page 35: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 -0.0006414

2 lam2 0.87495 -0.18254

3 lam3 0.72456 -0.0000787

4 _Parm1 0.17144 0.03574

5 _Parm2 0.23105 0.19740

6 _Parm3 0.26848 0.00184

7 _Add01 4.43522 -0.07737

8 _Add02 7.16990 -0.05046

9 _Add03 0.53613 -0.35677

10 _Add04 1.43723 0.10297

11 _Add05 0.96724 -0.09405

12 _Add06 1.54578 0.07587

13 _Add07 0.44156 -0.32954

14 _Add08 0.88661 -0.05917

15 _Add09 0.72424 -0.13889

16 _Add10 1.01554 -0.01546

17 _Add11 1.20242 0.02992

18 _Add12 0.84194 -0.02756

19 _Add13 1.28831 -0.00532

Page 36: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add14 1.09709 0.01769

21 _Add15 0.09057 -0.14819

22 _Add16 1.00923 0.01286

23 _Add17 4.84064 0.12916

Value of Objective Function = 0.4751696982

Page 37: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 23

Functions (Observations) 78

Optimization Start

Active Constraints 0 Objective Function 0.4751696982

Max Abs Gradient Element 0.3567744922 Radius 1.5014336492

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 * 0 4 0 0.32306 0.1521 0.1768 111E-16

0.870

2 * 0 6 0 0.31999 0.00306 0.00789 111E-16

1.011

3 * 0 8 0 0.31983 0.000159

0.00810 111E-16

1.164

4 * 0 10 0 0.31982 0.000015

0.00116 111E-16

1.293

5 * 0 12 0 0.31982 1.757E-6

0.000665

111E-16

1.338

6 * 0 14 0 0.31982 2.174E-7

0.000183

111E-16

1.350

7 * 0 16 0 0.31982 2.723E- 0.00007 111E- 1.353

Page 38: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

8 4 16

8 * 0 18 0 0.31982 3.421E-9

0.000024

111E-16

1.354

9 * 0 20 0 0.31982 4.3E-10 8.98E-6 111E-16

1.354

Optimization Results

Iterations 9 Function Calls 23

Jacobian Calls 11 Active Constraints 0

Objective Function 0.3198177803 Max Abs Gradient Element 8.9797268E-6

Lambda 1.110223E-14 Actual Over Pred Change 1.3544493549

Radius 1729.6515639

Convergence criterion (GCONV=1E-8) satisfied.

Note: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.

Warning: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.

Note: Covariance matrix for the estimates is not full rank.

Note: The variance of some parameter estimates is zero or some parameter estimates are linearly related to other parameter estimates as shown in the following equations:

_Add17

= 27.471302

+ 0.158778

* lam1

+ 0.139907

* lam2

+ 0.116072

* lam3

- 1.413695

* _Add01

- 2.294513

* _Add02

- 0.025494

* _Add15

- 0.312201

* _Add16

Page 39: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)
Page 40: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 12

N Moments 78

N Parameters 23

N Active Constraints 0

Baseline Model Function Value 8.0815

Baseline Model Chi-Square 1834.4924

Baseline Model Chi-Square DF 66

Pr > Baseline Model Chi-Square <.0001

Absolute Index Fit Function 0.3198

Chi-Square 72.5986

Chi-Square DF 55

Pr > Chi-Square 0.0561

Z-Test of Wilson & Hilferty 1.5889

Hoelter Critical N 230

Root Mean Square Residual (RMSR) 0.1761

Standardized RMSR (SRMSR) 0.0458

Goodness of Fit Index (GFI) 0.9493

Parsimony Index Adjusted GFI (AGFI) 0.9280

Parsimonious GFI 0.7910

RMSEA Estimate 0.0375

RMSEA Lower 90% Confidence Limit 0.0000

RMSEA Upper 90% Confidence Limit 0.0592

Probability of Close Fit 0.8104

ECVI Estimate 0.5348

Page 41: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.4720

ECVI Upper 90% Confidence Limit 0.6526

Akaike Information Criterion 118.5986

Bozdogan CAIC 220.4736

Schwarz Bayesian Criterion 197.4736

McDonald Centrality 0.9621

Incremental Index

Bentler Comparative Fit Index 0.9900

Bentler-Bonett NFI 0.9604

Bentler-Bonett Non-normed Index 0.9881

Bollen Normed Index Rho1 0.9525

Bollen Non-normed Index Delta2 0.9901

James et al. Parsimonious NFI 0.8004

Page 42: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter Estimate

StandardError

t Value

attitude --->

X1 lam1 0.99298 0.04915 20.20360

attitude --->

X2 lam2 0.87497 0.04331 20.20120

attitude --->

X3 lam3 0.72591 0.03780 19.20429

subnorms --->

X4 lam1 0.99298 0.04915 20.20360

subnorms --->

X5 lam2 0.87497 0.04331 20.20120

subnorms --->

X6 lam3 0.72591 0.03780 19.20429

intention --->

X7 lam1 0.99298 0.04915 20.20360

intention --->

X8 lam2 0.87497 0.04331 20.20120

intention --->

X9 lam3 0.72591 0.03780 19.20429

beh --->

X13 lam1 0.99298 0.04915 20.20360

beh --->

X14 lam2 0.87497 0.04331 20.20120

beh --->

X15 lam3 0.72591 0.03780 19.20429

attitude --->

intention _Parm1 0.19491 0.05446 3.57881

subnorms --->

intention _Parm2 0.21164 0.04212 5.02501

intention --- beh _Parm3 0.27030 0.08581 3.15016

Page 43: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

PATH List

Path Parameter Estimate

StandardError

t Value

>

Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 4.42058 0.26050 16.96990

subnorms _Add02 7.17487 0.21277 33.72097

Error X1 _Add03 0.96696 0.14030 6.89207

X2 _Add04 1.03694 0.13055 7.94261

X3 _Add05 0.99635 0.11402 8.73828

X4 _Add06 1.11849 0.15461 7.23416

X5 _Add07 0.72892 0.11041 6.60204

X6 _Add08 0.93287 0.10954 8.51614

X7 _Add09 0.80416 0.09486 8.47695

X8 _Add10 1.03433 0.10812 9.56629

X9 _Add11 1.14329 0.11357 10.06694

X13 _Add12 0.90792 0.13571 6.69022

X14 _Add13 1.28909 0.15103 8.53510

X15 _Add14 1.06081 0.11932 8.89038

intention _Add15 0.07972 0.05739 1.38902

beh _Add16 0.97624 0.16629 5.87082

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate

StandardError

t Value

subnorms

attitude

_Add17 4.81665 0.12501 38.53129

Page 44: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 0.96696 5.32572 0.8184

X13 0.90792 1.94013 0.5320

X14 1.28909 2.09054 0.3834

X15 1.06081 1.61244 0.3421

X2 1.03694 4.42123 0.7655

X3 0.99635 3.32573 0.7004

X4 1.11849 8.19303 0.8635

X5 0.72892 6.22181 0.8828

X6 0.93287 4.71359 0.8021

X7 0.80416 1.75706 0.5423

X8 1.03433 1.77420 0.4170

X9 1.14329 1.65253 0.3082

beh 0.97624 1.04685 0.0675

intention 0.07972 0.96641 0.9175

Page 45: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter Estimate

StandardError

t Value

attitude --->

X1 lam1 0.90467 0.01550 58.35350

attitude --->

X2 lam2 0.87491 0.01769 49.44716

attitude --->

X3 lam3 0.83691 0.02076 40.32058

subnorms --->

X4 lam1 0.92924 0.01139 81.58953

subnorms --->

X5 lam2 0.93960 0.01050 89.46758

subnorms --->

X6 lam3 0.89559 0.01464 61.17299

intention --->

X7 lam1 0.73643 0.03369 21.86136

intention --->

X8 lam2 0.64577 0.03507 18.41129

intention --->

X9 lam3 0.55512 0.03500 15.85921

beh --->

X13 lam1 0.72941 0.03929 18.56413

beh --->

X14 lam2 0.61917 0.03716 16.66423

beh --->

X15 lam3 0.58490 0.03697 15.81994

attitude --->

intention _Parm1 0.41685 0.11486 3.62938

subnorms --->

intention _Parm2 0.57667 0.11209 5.14493

intention --- beh _Parm3 0.25971 0.07946 3.26850

Page 46: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for PATH List

Path Parameter Estimate

StandardError

t Value

>

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 1.00000

subnorms _Add02 1.00000

Error X1 _Add03 0.18156 0.02805 6.47261

X2 _Add04 0.23454 0.03096 7.57529

X3 _Add05 0.29959 0.03474 8.62323

X4 _Add06 0.13652 0.02117 6.44967

X5 _Add07 0.11715 0.01974 5.93624

X6 _Add08 0.19791 0.02622 7.54700

X7 _Add09 0.45767 0.04962 9.22448

X8 _Add10 0.58299 0.04530 12.86958

X9 _Add11 0.69184 0.03886 17.80265

X13 _Add12 0.46797 0.05732 8.16433

X14 _Add13 0.61663 0.04601 13.40190

X15 _Add14 0.65789 0.04325 15.21125

intention _Add15 0.08249 0.05547 1.48711

beh _Add16 0.93255 0.04127 22.59477

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add17 0.85526 0.02363 36.19076

Page 47: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simulation #4

Tested: Model 3

Data Set: Model 3_Behavior (ATT, PB, Intention in relation to Behavior)

SAS SYNTAX:

proc calis data=SASUSER.SEM_BEHDATA; path attitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, pastbeh ---> X10 X11 X12 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention beh, pastbeh---> intention beh, subnorms ---> intention, intention ---> beh; run;

SAS LOG:

1 proc calis data=SASUSER.SEM_BEHDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 pastbeh ---> X10 X11 X12 = lam1 lam2 lam3,7 beh ---> X13 X14 X15 = lam1 lam2 lam3,8 attitude ---> intention beh,9 pastbeh---> intention beh,10 subnorms ---> intention,11 intention ---> beh;12 run;

NOTE: Convergence criterion (GCONV=1E-8) satisfied.NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.WARNING: The estimated error variance for variable beh is negative.WARNING: Although all predicted variances for the latent variables are positive, the corresponding predicted covariance matrix is not positive definite. It has one negative eigenvalue.NOTE: PROCEDURE CALIS used (Total process time): real time 1.16 seconds cpu time 0.82 seconds

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_BEHDATA

Page 48: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Modeling Information

N Records Read

228

N Records Used 228

N Obs 228

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X10 X11 X12 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Latent beh intention

Exogenous Manifest

Latent attitude pastbeh subnorms

Number of Endogenous Variables = 17Number of Exogenous Variables = 3

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 0.08463 3.49124

X2 X2 -0.00553 3.28606

X3 X3 -0.44982 2.72201

X4 X4 -4.65034 7.64805

X5 X5 -4.26140 6.60533

X6 X6 4.05727 5.73898

X7 X7 2.00110 2.86434

X8 X8 -0.93788 2.66446

X9 X9 -3.42226 2.53520

X10

X10 0.00404 2.52169

Page 49: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simple Statistics

Variable Mean Std Dev

X11

X11 -0.22038 2.46534

X12

X12 1.88980 2.23842

X13

X13 -0.10176 1.29754

X14

X14 1.34148 1.23186

X15

X15 -0.03961 1.27580

Page 50: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 0.36281

2 lam2 0.88856 12.45473

3 lam3 0.77353 -17.20553

4 _Parm1 0.05729 0.91067

5 _Parm2 0.28379 0.17005

6 _Parm3 0.06844 0.94429

7 _Parm4 -0.08674 -0.08309

8 _Parm5 0.33393 5.45830

9 _Parm6 0.05288 0.10275

10 _Add01 11.31996 -0.15037

11 _Add02 55.05647 -0.01499

12 _Add03 6.24686 -0.28895

13 _Add04 0.86877 -0.22635

14 _Add05 1.86074 0.12035

15 _Add06 0.63606 -0.51062

16 _Add07 3.43625 0.10035

17 _Add08 0.16154 -77.60188

18 _Add09 0.01000 -122.58948

19 _Add10 0.27433 -7.13510

Page 51: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add11 0.83829 -0.26359

21 _Add12 1.68223 0.19783

22 _Add13 0.11203 -1.74138

23 _Add14 1.14578 -0.10634

24 _Add15 1.27273 -0.03011

25 _Add16 0.93906 -0.06054

26 _Add17 0.92964 0.00208

27 _Add18 1.18217 0.11600

28 _Add19 0.07663 -7.07095

29 _Add20 0.01000 0.21939

30 _Add21 22.47169 0.02911

31 _Add22 7.68976 0.32969

32 _Add23 15.92070 0.04665

Value of Objective Function = 14.240952318

Page 52: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 32

Functions (Observations) 120

Optimization Start

Active Constraints 0 Objective Function 14.240952318

Max Abs Gradient Element 122.58948369 Radius 1316.9455338

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 * 0 4 0 0.63309 13.6079 2.9088 111E-16

0.127

2 * 0 6 0 0.57576 0.0573 0.3575 111E-16

0.776

3 0 8 0 0.57192 0.00384 0.1853 0 0.853

4 * 0 10 0 0.57169 0.000223

0.0193 111E-16

0.832

5 * 0 12 0 0.57168 0.000014

0.0112 111E-16

0.854

6 0 14 0 0.57168 8.25E-7 0.000968

0 0.872

7 * 0 16 0 0.57168 5.14E-8 0.000695

111E-16

0.897

Page 53: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

8 * 0 18 0 0.57168 3.263E-9

0.000048

111E-16

0.924

9 0 20 0 0.57168 2.11E-10

0.000044

0 0.954

Optimization Results

Iterations 9 Function Calls 23

Jacobian Calls 11 Active Constraints 0

Objective Function 0.5716787581 Max Abs Gradient Element 0.0000440509

Lambda 0 Actual Over Pred Change 0.9537986891

Radius 0.0070486779

Convergence criterion (GCONV=1E-8) satisfied.

Note: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.

Warning: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.

Note: Covariance matrix for the estimates is not full rank.

Note: The variance of some parameter estimates is zero or some parameter estimates are linearly related to other parameter estimates as shown in the following equations:

_Add

= 1803.276

+0.166

*lam1

+0.147

*lam2

+0.127

*lam3

- 22.27

*_Add

- 2.459

*_Add

- 0.006

*_Add

+0.029

*_Add

- 9.007

*_Add

- 3.062

*_Add

- 6.362

*_Add

Page 54: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

01

353

895

046

770

3729

02

546

03

467

19

204

20

129

21

557

22

338

23

Page 55: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 15

N Moments 120

N Parameters 32

N Active Constraints 0

Baseline Model Function Value 23.2661

Baseline Model Chi-Square 5281.3943

Baseline Model Chi-Square DF 105

Pr > Baseline Model Chi-Square <.0001

Absolute Index Fit Function 0.5717

Chi-Square 129.7711

Chi-Square DF 88

Pr > Chi-Square 0.0025

Z-Test of Wilson & Hilferty 2.8011

Hoelter Critical N 194

Root Mean Square Residual (RMSR) 0.5067

Standardized RMSR (SRMSR) 0.0560

Goodness of Fit Index (GFI) 0.9258

Parsimony Index Adjusted GFI (AGFI) 0.8989

Parsimonious GFI 0.7759

RMSEA Estimate 0.0457

RMSEA Lower 90% Confidence Limit 0.0276

RMSEA Upper 90% Confidence Limit 0.0618

Probability of Close Fit 0.6500

ECVI Estimate 0.8750

Page 56: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.7573

ECVI Upper 90% Confidence Limit 1.0309

Akaike Information Criterion 193.7711

Bozdogan CAIC 335.5101

Schwarz Bayesian Criterion 303.5101

McDonald Centrality 0.9125

Incremental Index

Bentler Comparative Fit Index 0.9919

Bentler-Bonett NFI 0.9754

Bentler-Bonett Non-normed Index 0.9904

Bollen Normed Index Rho1 0.9707

Bollen Non-normed Index Delta2 0.9920

James et al. Parsimonious NFI 0.8175

Page 57: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter Estimate

StandardError

t Value

attitude --->

X1 lam1 0.94411 0.04452 21.20833

attitude --->

X2 lam2 0.83182 0.03957 21.01906

attitude --->

X3 lam3 0.72278 0.03446 20.97371

subnorms --->

X4 lam1 0.94411 0.04452 21.20833

subnorms --->

X5 lam2 0.83182 0.03957 21.01906

subnorms --->

X6 lam3 0.72278 0.03446 20.97371

intention --->

X7 lam1 0.94411 0.04452 21.20833

intention --->

X8 lam2 0.83182 0.03957 21.01906

intention --->

X9 lam3 0.72278 0.03446 20.97371

pastbeh --->

X10 lam1 0.94411 0.04452 21.20833

pastbeh --->

X11 lam2 0.83182 0.03957 21.01906

pastbeh --->

X12 lam3 0.72278 0.03446 20.97371

beh --->

X13 lam1 0.94411 0.04452 21.20833

beh --->

X14 lam2 0.83182 0.03957 21.01906

beh --- X15 lam3 0.72278 0.03446 20.97371

Page 58: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

PATH List

Path Parameter Estimate

StandardError

t Value

>

attitude --->

intention _Parm1 0.06746 0.04974 1.35613

attitude --->

beh _Parm2 0.28658 0.04922 5.82215

pastbeh --->

intention _Parm3 0.08116 0.05882 1.37982

pastbeh --->

beh _Parm4 -0.08894 0.05576 -1.59493

subnorms --->

intention _Parm5 0.32410 0.01532 21.15536

intention --->

beh _Parm6 0.05082 0.04363 1.16463

Variance Parameters

VarianceType

Variable Parameter Estimate

StandardError

t Value

Exogenous

attitude _Add01 12.80340 0.73405 17.44209

subnorms _Add02 62.99993 0.60672 103.83612

pastbeh _Add03 6.95668 0.48362 14.38449

Error X1 _Add04 1.12245 0.15343 7.31578

X2 _Add05 1.21204 0.14641 8.27854

X3 _Add06 1.01754 0.11933 8.52693

X4 _Add07 0.79370 0.14471 5.48486

X5 _Add08 1.26576 0.15723 8.05035

X6 _Add09 1.02833 0.12459 8.25387

X7 _Add10 1.01916 0.11062 9.21336

X8 _Add11 0.91068 0.09601 9.48540

X9 _Add12 1.05184 0.10487 10.03019

X10 _Add13 0.78918 0.11666 6.76460

Page 59: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Variance Parameters

VarianceType

Variable Parameter Estimate

StandardError

t Value

X11 _Add14 0.91937 0.11390 8.07138

X12 _Add15 0.96620 0.10945 8.82809

X13 _Add16 1.06074 0.10911 9.72145

X14 _Add17 0.98646 0.09856 10.00863

X15 _Add18 1.03758 0.10015 10.36060

intention _Add19 0.01829 0.06028 0.30345

beh _Add20 -0.08260 0.05478 -1.50789

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add21 25.47614 0.71109 35.82691

pastbeh attitude _Add22 8.66226 0.49140 17.62772

pastbeh subnorms _Add23 17.99550 0.66843 26.92206

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 1.12245 12.53462 0.9105

X10 0.78918 6.98994 0.8871

X11 0.91937 5.73292 0.8396

X12 0.96620 4.60047 0.7900

X13 1.06074 1.79911 0.4104

X14 0.98646 1.55965 0.3675

X15 1.03758 1.47034 0.2943

X2 1.21204 10.07113 0.8797

X3 1.01754 7.70621 0.8680

X4 0.79370 56.94800 0.9861

X5 1.26576 44.85745 0.9718

X6 1.02833 33.94037 0.9697

Page 60: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X7 1.01916 8.94791 0.8861

X8 0.91068 7.06564 0.8711

X9 1.05184 5.69888 0.8154

beh -0.08260 0.82839 .

intention 0.01829 8.89531 0.9979

Page 61: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter Estimate StandardError

t Value

attitude ---> X1 lam1 0.95418 0.00729 130.85963

attitude ---> X2 lam2 0.93790 0.00890 105.41700

attitude ---> X3 lam3 0.93164 0.00953 97.76019

subnorms

---> X4 lam1 0.99301 0.00143 693.47773

subnorms

---> X5 lam2 0.98579 0.00219 449.24742

subnorms

---> X6 lam3 0.98473 0.00232 424.56719

intention ---> X7 lam1 0.94133 0.00801 117.59053

intention ---> X8 lam2 0.93333 0.00884 105.55467

intention ---> X9 lam3 0.90301 0.01189 75.95131

pastbeh ---> X10 lam1 0.94186 0.00968 97.30359

pastbeh ---> X11 lam2 0.91632 0.01187 77.22318

pastbeh ---> X12 lam3 0.88881 0.01429 62.20646

beh ---> X13 lam1 0.64063 0.03826 16.74237

beh ---> X14 lam2 0.60623 0.03812 15.90372

beh ---> X15 lam3 0.54252 0.03707 14.63637

attitude ---> intention

_Parm1 0.08093 0.05969 1.35596

attitude ---> beh _Parm2 1.12666 0.18908 5.95859

pastbeh ---> intention

_Parm3 0.07177 0.05202 1.37974

pastbeh ---> beh _Parm4 -0.25773 0.16149 -1.59594

subnorms

---> intention

_Parm5 0.86251 0.03839 22.46427

Page 62: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for PATH List

Path Parameter Estimate StandardError

t Value

intention ---> beh _Parm6 0.16653 0.14293 1.16508

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 1.00000

subnorms _Add02 1.00000

pastbeh _Add03 1.00000

Error X1 _Add04 0.08955 0.01391 6.43537

X2 _Add05 0.12035 0.01669 7.21124

X3 _Add06 0.13204 0.01776 7.43606

X4 _Add07 0.01394 0.00284 4.90091

X5 _Add08 0.02822 0.00433 6.52233

X6 _Add09 0.03030 0.00457 6.63276

X7 _Add10 0.11390 0.01507 7.55757

X8 _Add11 0.12889 0.01651 7.80887

X9 _Add12 0.18457 0.02147 8.59568

X10 _Add13 0.11290 0.01823 6.19196

X11 _Add14 0.16037 0.02175 7.37463

X12 _Add15 0.21002 0.02540 8.26908

X13 _Add16 0.58959 0.04903 12.02596

X14 _Add17 0.63249 0.04622 13.68527

X15 _Add18 0.70567 0.04022 17.54591

intention _Add19 0.00206 0.00677 0.30397

beh _Add20 -0.09971 0.07259 -1.37360

Page 63: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add21 0.89702 0.01471 60.96711

pastbeh attitude _Add22 0.91784 0.01436 63.93393

pastbeh subnorms _Add23 0.85959 0.01971 43.61092

Simulation #5

Tested: Model 2

Data Set: Model 3_Behavior (ATT, PB, Intention in relation to Behavior)

SAS SYNTAX:

proc calis data=SASUSER.SEM_BEHDATA; path attitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention beh, subnorms ---> intention, intention ---> beh;run;

SAS LOG:

1 proc calis data=SASUSER.SEM_BEHDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 beh ---> X13 X14 X15 = lam1 lam2 lam3,7 attitude ---> intention beh,8 subnorms ---> intention,9 intention ---> beh;10 run;

NOTE: Convergence criterion (GCONV=1E-8) satisfied.NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.WARNING: The estimated error variance for variable beh is negative.WARNING: Although all predicted variances for the latent variables are positive, the corresponding predicted covariance matrix is not positive definite. It has one negative eigenvalue.NOTE: PROCEDURE CALIS used (Total process time): real time 1.11 seconds cpu time 0.76 seconds

Page 64: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_BEHDATA

N Records Read

228

N Records Used 228

N Obs 228

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Latent beh intention

Exogenous Manifest

Latent attitude subnorms

Number of Endogenous Variables = 14Number of Exogenous Variables = 2

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 0.08463 3.49124

X2 X2 -0.00553 3.28606

X3 X3 -0.44982 2.72201

X4 X4 -4.65034 7.64805

X5 X5 -4.26140 6.60533

X6 X6 4.05727 5.73898

Page 65: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simple Statistics

Variable Mean Std Dev

X7 X7 2.00110 2.86434

X8 X8 -0.93788 2.66446

X9 X9 -3.42226 2.53520

X13

X13 -0.10176 1.29754

X14

X14 1.34148 1.23186

X15

X15 -0.03961 1.27580

Page 66: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 0.05099

2 lam2 0.88258 -0.71220

3 lam3 0.76698 -1.00534

4 _Parm1 0.09720 1.79477

5 _Parm2 0.23575 0.13456

6 _Parm3 0.33712 4.21131

7 _Parm4 0.03835 0.00696

8 _Add01 11.42750 -0.01591

9 _Add02 55.58316 -0.00348

10 _Add03 0.76123 -0.27455

11 _Add04 1.89673 0.11392

12 _Add05 0.68699 -0.41200

13 _Add06 2.90956 0.16265

14 _Add07 0.33377 -4.58306

15 _Add08 0.23842 -6.76093

16 _Add09 0.20017 -8.93301

17 _Add10 0.86443 -0.18205

18 _Add11 1.71860 0.19553

19 _Add12 0.93214 -0.06896

Page 67: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add13 0.93212 0.0004969

21 _Add14 1.18560 0.11979

22 _Add15 0.09373 -6.99976

23 _Add16 0.01000 0.20968

24 _Add17 22.66725 0.01566

Value of Objective Function = 3.8871557782

Page 68: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 24

Functions (Observations) 78

Optimization Start

Active Constraints 0 Objective Function 3.8871557782

Max Abs Gradient Element 8.9330099601 Radius 97.501600999

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 * 0 4 0 0.39783 3.4893 1.3617 111E-16

0.404

2 * 0 6 0 0.36513 0.0327 0.0915 111E-16

0.974

3 * 0 8 0 0.36348 0.00165 0.0666 111E-16

1.012

4 * 0 10 0 0.36341 0.000072

0.00598 111E-16

1.033

5 * 0 12 0 0.36340 3.573E-6

0.00307 111E-16

1.064

6 * 0 14 0 0.36340 1.811E-7

0.000283

111E-16

1.091

7 * 0 16 0 0.36340 9.36E-9 0.00014 111E- 1.116

Page 69: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

0 16

8 0 18 0 0.36340 4.9E-10 0.000014

0 1.137

Optimization Results

Iterations 8 Function Calls 21

Jacobian Calls 10 Active Constraints 0

Objective Function 0.3634029965 Max Abs Gradient Element 0.0000136979

Lambda 0 Actual Over Pred Change 1.1365563738

Radius 0.015509586

Convergence criterion (GCONV=1E-8) satisfied.

Note: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.

Warning: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.

Note: Covariance matrix for the estimates is not full rank.

Note: The variance of some parameter estimates is zero or some parameter estimates are linearly related to other parameter estimates as shown in the following equations:

lam2

= -88.642173

- 0.012817

* lam1

- 0.009719

* lam3

+ 0.277600

* _Add01

+ 1.366098

* _Add02

+ 0.000498

* _Add15

- 0.001615

* _Add16

+ 0.552380

* _Add17

Page 70: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 12

N Moments 78

N Parameters 24

N Active Constraints 0

Baseline Model Function Value 18.7765

Baseline Model Chi-Square 4262.2653

Baseline Model Chi-Square DF 66

Pr > Baseline Model Chi-Square <.0001

Absolute Index Fit Function 0.3634

Chi-Square 82.4925

Chi-Square DF 54

Pr > Chi-Square 0.0075

Z-Test of Wilson & Hilferty 2.4290

Hoelter Critical N 199

Root Mean Square Residual (RMSR) 0.5319

Standardized RMSR (SRMSR) 0.0558

Goodness of Fit Index (GFI) 0.9406

Parsimony Index Adjusted GFI (AGFI) 0.9143

Parsimonious GFI 0.7696

RMSEA Estimate 0.0482

RMSEA Lower 90% Confidence Limit 0.0254

RMSEA Upper 90% Confidence Limit 0.0682

Probability of Close Fit 0.5347

ECVI Estimate 0.5877

Page 71: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.4965

ECVI Upper 90% Confidence Limit 0.7163

Akaike Information Criterion 130.4925

Bozdogan CAIC 236.7968

Schwarz Bayesian Criterion 212.7968

McDonald Centrality 0.9394

Incremental Index

Bentler Comparative Fit Index 0.9932

Bentler-Bonett NFI 0.9806

Bentler-Bonett Non-normed Index 0.9917

Bollen Normed Index Rho1 0.9763

Bollen Non-normed Index Delta2 0.9932

James et al. Parsimonious NFI 0.8023

Page 72: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter Estimate

StandardError

t Value

attitude --->

X1 lam1 1.02096 0.04824 21.16326

attitude --->

X2 lam2 0.89105 0.04256 20.93433

attitude --->

X3 lam3 0.77418 0.03705 20.89547

subnorms --->

X4 lam1 1.02096 0.04824 21.16326

subnorms --->

X5 lam2 0.89105 0.04256 20.93433

subnorms --->

X6 lam3 0.77418 0.03705 20.89547

intention --->

X7 lam1 1.02096 0.04824 21.16326

intention --->

X8 lam2 0.89105 0.04256 20.93433

intention --->

X9 lam3 0.77418 0.03705 20.89547

beh --->

X13 lam1 1.02096 0.04824 21.16326

beh --->

X14 lam2 0.89105 0.04256 20.93433

beh --->

X15 lam3 0.77418 0.03705 20.89547

attitude --->

intention _Parm1 0.11243 0.03462 3.24767

attitude --->

beh _Parm2 0.23878 0.03610 6.61452

subnorms --- intention _Parm3 0.32927 0.01541 21.36184

Page 73: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

PATH List

Path Parameter Estimate

StandardError

t Value

>

intention --->

beh _Parm4 0.03464 0.04262 0.81269

Variance Parameters

VarianceType

Variable Parameter Estimate

StandardError

t Value

Exogenous

attitude _Add01 11.05675 0.65438 16.89647

subnorms _Add02 54.41143 0.38422 141.61406

Error X1 _Add03 1.14034 0.16295 6.99802

X2 _Add04 1.24634 0.15358 8.11515

X3 _Add05 0.98082 0.11929 8.22242

X4 _Add06 0.75622 0.14306 5.28609

X5 _Add07 1.27384 0.15688 8.11958

X6 _Add08 1.03241 0.12418 8.31376

X7 _Add09 1.02285 0.11166 9.16004

X8 _Add10 0.91879 0.09689 9.48237

X9 _Add11 1.05148 0.10499 10.01526

X13 _Add12 1.06234 0.10957 9.69559

X14 _Add13 0.98983 0.09881 10.01711

X15 _Add14 1.03485 0.09989 10.35960

intention _Add15 0.01983 0.05296 0.37451

beh _Add16 -0.06434 0.04680 -1.37480

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate

StandardError

t Value

subnorms

attitude

_Add17 22.00121 0.65249 33.71885

Page 74: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 1.14034 12.66547 0.9100

X13 1.06234 1.80832 0.4125

X14 0.98983 1.55804 0.3647

X15 1.03485 1.46379 0.2930

X2 1.24634 10.02501 0.8757

X3 0.98082 7.60780 0.8711

X4 0.75622 57.47256 0.9868

X5 1.27384 44.47460 0.9714

X6 1.03241 33.64447 0.9693

X7 1.02285 9.03657 0.8868

X8 0.91879 7.02283 0.8692

X9 1.05148 5.65939 0.8142

beh -0.06434 0.71566 .

intention 0.01983 7.68805 0.9974

Page 75: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter Estimate StandardError

t Value

attitude ---> X1 lam1 0.95392 0.00756 126.24040

attitude ---> X2 lam2 0.93578 0.00929 100.73376

attitude ---> X3 lam3 0.93331 0.00954 97.87219

subnorms

---> X4 lam1 0.99340 0.00139 713.20903

subnorms

---> X5 lam2 0.98557 0.00222 444.51270

subnorms

---> X6 lam3 0.98454 0.00234 420.67693

intention ---> X7 lam1 0.94171 0.00799 117.88479

intention ---> X8 lam2 0.93229 0.00897 103.90016

intention ---> X9 lam3 0.90233 0.01198 75.32850

beh ---> X13 lam1 0.64228 0.03832 16.76236

beh ---> X14 lam2 0.60390 0.03815 15.82949

beh ---> X15 lam3 0.54133 0.03709 14.59602

attitude ---> intention

_Parm1 0.13484 0.04155 3.24523

attitude ---> beh _Parm2 0.93853 0.13766 6.81753

subnorms

---> intention

_Parm3 0.87598 0.03851 22.74563

intention ---> beh _Parm4 0.11353 0.13966 0.81287

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 1.00000

Page 76: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

subnorms _Add02 1.00000

Error X1 _Add03 0.09004 0.01442 6.24535

X2 _Add04 0.12432 0.01739 7.15076

X3 _Add05 0.12892 0.01780 7.24276

X4 _Add06 0.01316 0.00277 4.75473

X5 _Add07 0.02864 0.00437 6.55356

X6 _Add08 0.03069 0.00461 6.65874

X7 _Add09 0.11319 0.01505 7.52324

X8 _Add10 0.13083 0.01673 7.81961

X9 _Add11 0.18579 0.02162 8.59459

X13 _Add12 0.58747 0.04922 11.93551

X14 _Add13 0.63530 0.04608 13.78749

X15 _Add14 0.70697 0.04015 17.60692

intention _Add15 0.00258 0.00687 0.37530

beh _Add16 -0.08990 0.07135 -1.25997

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add17 0.89699 0.01472 60.93434

Page 77: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simulation #6

Tested: Model 1

Data Set: Model 3_Behavior (ATT, PB, Intention in relation to Behavior)

SAS SYNTAX:

proc calis data=SASUSER.SEM_BEHDATA; pathattitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention, subnorms ---> intention, intention ---> beh;run;

SAS LOG:

1 proc calis data=SASUSER.SEM_BEHDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 beh ---> X13 X14 X15 = lam1 lam2 lam3,7 attitude ---> intention,8 subnorms ---> intention,9 intention ---> beh;10 run;

NOTE: Convergence criterion (GCONV=1E-8) satisfied.NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.WARNING: The estimated error variance for variable intention is negative.WARNING: Although all predicted variances for the latent variables are positive, the corresponding predicted covariance matrix is not positive definite. It has one negative eigenvalue.NOTE: PROCEDURE CALIS used (Total process time): real time 1.16 seconds cpu time 0.73 seconds

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_BEHDATA

N Records Read

228

Page 78: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Modeling Information

N Records Used 228

N Obs 228

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Latent beh intention

Exogenous Manifest

Latent attitude subnorms

Number of Endogenous Variables = 14Number of Exogenous Variables = 2

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 0.08463 3.49124

X2 X2 -0.00553 3.28606

X3 X3 -0.44982 2.72201

X4 X4 -4.65034 7.64805

X5 X5 -4.26140 6.60533

X6 X6 4.05727 5.73898

X7 X7 2.00110 2.86434

X8 X8 -0.93788 2.66446

X9 X9 -3.42226 2.53520

X13

X13 -0.10176 1.29754

X1 X14 1.34148 1.23186

Page 79: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simple Statistics

Variable Mean Std Dev

4

X15

X15 -0.03961 1.27580

Page 80: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 0.02184

2 lam2 0.88258 -0.65891

3 lam3 0.76698 -1.00265

4 _Parm1 0.09720 1.55402

5 _Parm2 0.33712 4.28227

6 _Parm3 0.29918 0.08160

7 _Add01 11.42750 -0.01855

8 _Add02 55.58316 -0.00387

9 _Add03 0.76123 -0.27502

10 _Add04 1.89673 0.10599

11 _Add05 0.68699 -0.38154

12 _Add06 2.90956 0.16250

13 _Add07 0.33377 -4.55207

14 _Add08 0.23842 -6.79937

15 _Add09 0.20017 -9.05307

16 _Add10 0.86443 -0.18029

17 _Add11 1.71860 0.19560

18 _Add12 0.93214 -0.04391

19 _Add13 0.93212 -0.08743

Page 81: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add14 1.18560 0.09489

21 _Add15 0.09373 -6.80858

22 _Add16 0.06163 -0.00223

23 _Add17 22.66725 0.01770

Value of Objective Function = 4.0659704694

Page 82: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 23

Functions (Observations) 78

Optimization Start

Active Constraints 0 Objective Function 4.0659704694

Max Abs Gradient Element 9.0530710437 Radius 99.053838136

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 * 0 4 0 0.58507 3.4809 1.3512 111E-16

0.401

2 * 0 6 0 0.53391 0.0512 0.1266 111E-16

0.971

3 * 0 8 0 0.52943 0.00448 0.0922 111E-16

0.997

4 * 0 10 0 0.52915 0.000284

0.0102 111E-16

0.984

5 * 0 12 0 0.52912 0.000022

0.00702 111E-16

0.981

6 * 0 14 0 0.52912 1.73E-6 0.000872

111E-16

0.979

7 * 0 16 0 0.52912 1.37E-7 0.00055 111E- 0.976

Page 83: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

7 16

8 * 0 18 0 0.52912 1.087E-8

0.000072

111E-16

0.974

9 * 0 20 0 0.52912 8.63E-10

0.000044

111E-16

0.972

Optimization Results

Iterations 9 Function Calls 23

Jacobian Calls 11 Active Constraints 0

Objective Function 0.5291221419 Max Abs Gradient Element 0.0000443914

Lambda 1.110223E-14 Actual Over Pred Change 0.9719668535

Radius 57055.010766

Convergence criterion (GCONV=1E-8) satisfied.

Note: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.

Warning: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.

Note: Covariance matrix for the estimates is not full rank.

Note: The variance of some parameter estimates is zero or some parameter estimates are linearly related to other parameter estimates as shown in the following equations:

_Add02

= 1878.002217

+ 1.53227

* lam1

+ 1.33565

* lam2

+ 1.16154

* lam3

- 33.249347

* _Add01

+ 0.01345

* _Add15

- 0.06759

* _Add16

- 66.203578

* _Add17

Page 84: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

2 1 3 3 5

Page 85: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 12

N Moments 78

N Parameters 23

N Active Constraints 0

Baseline Model Function Value 18.7765

Baseline Model Chi-Square 4262.2653

Baseline Model Chi-Square DF 66

Pr > Baseline Model Chi-Square <.0001

Absolute Index Fit Function 0.5291

Chi-Square 120.1107

Chi-Square DF 55

Pr > Chi-Square <.0001

Z-Test of Wilson & Hilferty 4.7423

Hoelter Critical N 139

Root Mean Square Residual (RMSR) 0.5419

Standardized RMSR (SRMSR) 0.0607

Goodness of Fit Index (GFI) 0.9162

Parsimony Index Adjusted GFI (AGFI) 0.8811

Parsimonious GFI 0.7635

RMSEA Estimate 0.0722

RMSEA Lower 90% Confidence Limit 0.0546

RMSEA Upper 90% Confidence Limit 0.0898

Probability of Close Fit 0.0206

ECVI Estimate 0.7441

Page 86: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.6196

ECVI Upper 90% Confidence Limit 0.9049

Akaike Information Criterion 166.1107

Bozdogan CAIC 267.9857

Schwarz Bayesian Criterion 244.9857

McDonald Centrality 0.8669

Incremental Index

Bentler Comparative Fit Index 0.9845

Bentler-Bonett NFI 0.9718

Bentler-Bonett Non-normed Index 0.9814

Bollen Normed Index Rho1 0.9662

Bollen Non-normed Index Delta2 0.9845

James et al. Parsimonious NFI 0.8098

Page 87: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter Estimate

StandardError

t Value

attitude --->

X1 lam1 1.02020 0.04821 21.16159

attitude --->

X2 lam2 0.88929 0.04248 20.93349

attitude --->

X3 lam3 0.77336 0.03701 20.89356

subnorms --->

X4 lam1 1.02020 0.04821 21.16159

subnorms --->

X5 lam2 0.88929 0.04248 20.93349

subnorms --->

X6 lam3 0.77336 0.03701 20.89356

intention --->

X7 lam1 1.02020 0.04821 21.16159

intention --->

X8 lam2 0.88929 0.04248 20.93349

intention --->

X9 lam3 0.77336 0.03701 20.89356

beh --->

X13 lam1 1.02020 0.04821 21.16159

beh --->

X14 lam2 0.88929 0.04248 20.93349

beh --->

X15 lam3 0.77336 0.03701 20.89356

attitude --->

intention _Parm1 0.16871 0.03323 5.07765

subnorms --->

intention _Parm2 0.30601 0.01482 20.65263

intention --- beh _Parm3 0.30105 0.01663 18.10657

Page 88: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

PATH List

Path Parameter Estimate

StandardError

t Value

>

Variance Parameters

VarianceType

Variable Parameter Estimate

StandardError

t Value

Exogenous

attitude _Add01 11.06884 0.65512 16.89600

subnorms _Add02 54.52913 0.38458 141.78801

Error X1 _Add03 1.13869 0.16694 6.82084

X2 _Add04 1.28985 0.15947 8.08844

X3 _Add05 0.95313 0.11878 8.02413

X4 _Add06 0.76001 0.14368 5.28963

X5 _Add07 1.26192 0.15600 8.08913

X6 _Add08 1.03772 0.12474 8.31943

X7 _Add09 1.06710 0.11251 9.48480

X8 _Add10 0.92693 0.09554 9.70170

X9 _Add11 1.06303 0.10475 10.14859

X13 _Add12 0.99417 0.10785 9.21852

X14 _Add13 1.03431 0.10568 9.78744

X15 _Add14 1.04740 0.10352 10.11829

intention _Add15 -0.00448 0.04707 -0.09514

beh _Add16 0.02250 0.05119 0.43963

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate

StandardError

t Value

subnorms

attitude

_Add17 22.03943 0.65362 33.71904

Page 89: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 1.13869 12.65920 0.9101

X13 0.99417 1.74321 0.4297

X14 1.03431 1.60344 0.3549

X15 1.04740 1.47783 0.2913

X2 1.28985 10.04343 0.8716

X3 0.95313 7.57332 0.8741

X4 0.76001 57.51425 0.9868

X5 1.26192 44.38526 0.9716

X6 1.03772 33.65117 0.9692

X7 1.06710 9.07335 0.8824

X8 0.92693 7.01028 0.8678

X9 1.06303 5.66377 0.8123

beh 0.02250 0.71967 0.9687

intention -0.00448 7.69236 .

Page 90: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter Estimate StandardError

t Value

attitude ---> X1 lam1 0.95397 0.00769 124.12312

attitude ---> X2 lam2 0.93358 0.00959 97.32278

attitude ---> X3 lam3 0.93496 0.00947 98.75318

subnorms

---> X4 lam1 0.99337 0.00140 710.50840

subnorms

---> X5 lam2 0.98568 0.00221 446.73248

subnorms

---> X6 lam3 0.98446 0.00235 418.75405

intention ---> X7 lam1 0.93936 0.00811 115.80896

intention ---> X8 lam2 0.93155 0.00894 104.18405

intention ---> X9 lam3 0.90128 0.01200 75.11723

beh ---> X13 lam1 0.65551 0.03862 16.97130

beh ---> X14 lam2 0.59577 0.03819 15.60053

beh ---> X15 lam3 0.53968 0.03708 14.55592

attitude ---> intention

_Parm1 0.20238 0.03988 5.07519

subnorms

---> intention

_Parm2 0.81474 0.03745 21.75328

intention ---> beh _Parm3 0.98424 0.03506 28.07583

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 1.00000

subnorms _Add02 1.00000

Page 91: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Error X1 _Add03 0.08995 0.01466 6.13419

X2 _Add04 0.12843 0.01791 7.17028

X3 _Add05 0.12585 0.01770 7.10891

X4 _Add06 0.01321 0.00278 4.75730

X5 _Add07 0.02843 0.00435 6.53637

X6 _Add08 0.03084 0.00463 6.66215

X7 _Add09 0.11761 0.01524 7.71772

X8 _Add10 0.13222 0.01666 7.93730

X9 _Add11 0.18769 0.02163 8.67811

X13 _Add12 0.57031 0.05064 11.26276

X14 _Add13 0.64505 0.04550 14.17562

X15 _Add14 0.70874 0.04002 17.71012

intention _Add15 -0.0005822 0.00612 -0.09510

beh _Add16 0.03127 0.06901 0.45311

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add17 0.89709 0.01471 60.96815

SIMULATION #7

Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Data Set: Random

SAS SYNTAX:

proc calis data=SASUSER.SEM_RANDOMDATA; path attitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3,

Page 92: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

pastbeh ---> X10 X11 X12 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention beh, pastbeh---> intention beh, subnorms ---> intention, intention ---> beh; run;

SAS LOG:

1 proc calis data=SASUSER.SEM_RANDOMDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 pastbeh ---> X10 X11 X12 = lam1 lam2 lam3,7 beh ---> X13 X14 X15 = lam1 lam2 lam3,8 attitude ---> intention beh,9 pastbeh---> intention beh,10 subnorms ---> intention,11 intention ---> beh;12 run;

ERROR: LEVMAR Optimization cannot be completed.NOTE: LEVMAR needs more than 50 iterations or 500 function calls.NOTE: Due to optimization failure, statistics in the Fit Summary table might not be appropriate.NOTE: Due to optimization failure, standard error estimates are not computed.WARNING: Number of zero or negative variances encountered when computing standardized results: 1. Standardization with these values was not done.NOTE: PROCEDURE CALIS used (Total process time): real time 1.26 seconds cpu time 0.82 seconds

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_RANDOMDATA

N Records Read

228

N Records Used 228

N Obs 228

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X10 X11 X12 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Page 93: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Variables in the Model

Latent beh intention

Exogenous Manifest

Latent attitude pastbeh subnorms

Number of Endogenous Variables = 17Number of Exogenous Variables = 3

Page 94: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 3.96942 1.67800

X2 X2 3.92882 1.68062

X3 X3 4.01325 1.74491

X4 X4 3.90928 1.74093

X5 X5 3.92321 1.80255

X6 X6 3.96071 1.68701

X7 X7 4.10400 1.78364

X8 X8 4.11832 1.75092

X9 X9 3.95523 1.67310

X10 X10

4.20442 1.66872

X11 X11

4.06377 1.73203

X12 X12

4.04980 1.69682

X13 X13

4.02703 1.70790

X14 X14

4.16736 1.81132

X15 X15

4.04412 1.73333

Page 95: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 -0.00814

2 lam2 0.10976 0.05704

3 lam3 0.35586 0.02582

4 _Parm1 0.00100 0.01048

5 _Parm2 0.01000 0.00494

6 _Parm3 0.00100 -0.00524

7 _Parm4 0.01000 0.0007097

8 _Parm5 0.00100 -0.01103

9 _Parm6 0.01000 -0.00147

10 _Add01 0.36572 0.0005640

11 _Add02 0.33939 -0.0001756

12 _Add03 0.10000 0.01353

13 _Add04 2.44997 -0.0003409

14 _Add05 2.82007 -0.00116

15 _Add06 2.99841 0.00140

16 _Add07 2.69147 0.00456

17 _Add08 3.24509 -0.0006077

18 _Add09 2.80304 -0.00348

19 _Add10 3.08138 -0.0005114

Page 96: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add11 3.06450 -0.0001119

21 _Add12 2.78659 -0.0004828

22 _Add13 2.68462 -0.0004177

23 _Add14 2.99872 -0.0002916

24 _Add15 2.86653 -0.0002666

25 _Add16 2.81694 -0.0002193

26 _Add17 3.27968 0.0000626

27 _Add18 2.99178 -0.0005133

28 _Add19 0.10000 0.01598

29 _Add20 0.10006 0.00955

30 _Add21 -0.29481 0.00135

31 _Add22 0 0.03727

32 _Add23 0 0.07304

Value of Objective Function = 0.4667643241

Page 97: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 32

Functions (Observations) 120

Optimization Start

Active Constraints 0 Objective Function 0.4667643241

Max Abs Gradient Element 0.0730430547 Radius 1

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 0 5 0 0.45034 0.0164 0.0290 1.598 0.677

2 * 0 7 0 0.44315 0.00719 0.0179 0.563 0.773

3 * 0 9 0 0.43869 0.00446 0.0146 0.0765 0.997

4 * 0 11 0 0.43305 0.00564 0.0163 0.0212 1.067

5 * 0 13 0 0.42660 0.00645 0.0225 0.00349

1.060

6 0 16 0 0.42350 0.00310 0.0107 0.0151 1.092

7 * 0 18 0 0.42057 0.00293 0.0217 0.00899

1.154

8 * 0 20 0 0.41995 0.000626

0.0610 0.00243

0.153

9 * 0 22 0 0.41603 0.00391 0.0915 0.0011 0.758

Page 98: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

0

10 0 25 0 0.41494 0.00110 0.00891 0.00332

0.942

11 * 0 27 0 0.41473 0.000208

0.0311 0.00139

0.743

12 0 29 0 0.41456 0.000167

0.0275 0.00091

0.862

13 0 32 0 0.41449 0.000072

0.00674 0.00126

0.967

14 0 34 0 0.41443 0.000060

0.0561 0.00050

0.635

15 * 0 36 0 0.41436 0.000069

0.0519 0.00034

0.780

16 * 0 39 0 0.41434 0.000025

0.000278

0.00135

1.005

17 0 41 0 0.41432 0.000017

0.0146 0.00064

0.905

18 * 0 43 0 0.41430 0.000023

0.0552 0.00025

0.687

19 0 45 0 0.41427 0.000025

0.0556 0.00021

0.770

20 0 48 0 0.41425 0.000020

0.0234 0.00023

0.927

21 * 0 50 0 0.41424 8.029E-6

0.1358 0.00010

0.293

22 0 52 0 0.41422 0.000027

0.1195 0.00008

0.722

23 0 54 0 0.41420 0.000019

0.1165 0.00007

0.722

24 0 56 0 0.41418 0.00001 0.1278 0.0000 0.686

Page 99: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

4 6

25 * 0 58 0 0.41417 0.000013

0.1387 0.00005

0.685

26 * 0 61 0 0.41416 8.877E-6

0.00644 0.00014

1.002

27 0 63 0 0.41416 5.601E-6

0.0620 0.00007

0.839

28 * 0 65 0 0.41415 9.789E-7

0.2714 0.00003

0.0744

29 0 67 0 0.41414 0.000016

0.0499 0.00006

0.966

30 0 69 0 0.41414 9.162E-7

0.2993 0.00003

0.0877

31 * 0 71 0 0.41412 0.000013

0.0573 0.00005

0.964

32 * 0 74 0 0.41412 3.038E-6

0.0372 0.00007

0.963

33 * 0 76 0 0.41412 2.806E-6

0.1791 0.00003

0.527

34 * 0 78 0 0.41411 5.169E-6

0.1790 0.00003

0.718

35 0 80 0 0.41411 4.299E-6

0.1955 0.00003

0.673

36 * 0 82 0 0.41410 5.166E-6

0.1412 0.00003

0.840

37 0 85 0 0.41410 2.076E-6

0.0167 0.00008

0.998

38 * 0 87 0 0.41410 2.158E-6

0.0570 0.00004

0.932

39 0 90 0 0.41410 1.931E- 0.0713 0.0000 0.902

Page 100: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

6 4

40 0 92 0 0.41409 2.778E-6

0.1900 0.00002

0.696

41 0 94 0 0.41409 3.632E-6

0.1857 0.00002

0.780

42 * 0 97 0 0.41409 2.573E-6

0.0440 0.00004

0.976

43 * 0 99 0 0.41409 8.566E-8

0.3885 0.00002

0.0256

44 0 101 0 0.41408 4.56E-6 0.0854 0.00003

0.961

45 0 103 0 0.41408 2.212E-6

0.2255 0.00002

0.700

46 * 0 105 0 0.41408 2.821E-6

0.2258 0.00001

0.763

47 0 108 0 0.41408 1.342E-6

0.0140 0.00007

1.000

48 * 0 110 0 0.41408 1.105E-6

0.0449 0.00003

0.965

49 0 112 0 0.41408 7.735E-7

0.3210 0.00002

0.365

50 0 114 0 0.41407 2.952E-6

0.1532 0.00002

0.903

Optimization Results

Iterations 50 Function Calls 117

Jacobian Calls 52 Active Constraints 0

Objective Function 0.4140728645 Max Abs Gradient Element 0.1532004038

Lambda 0.0000161945 Actual Over Pred Change 0.9025970153

Page 101: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization Results

Radius 0.6957418564

LEVMAR needs more than 50 iterations or 500 function calls.

LEVMAR Optimization cannot be completed.

Page 102: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 15

N Moments 120

N Parameters 32

N Active Constraints 0

Baseline Model Function Value 0.4801

Baseline Model Chi-Square 108.9878

Baseline Model Chi-Square DF 105

Pr > Baseline Model Chi-Square 0.3754

Absolute Index Fit Function 0.4141

Chi-Square 93.9945

Chi-Square DF 88

Pr > Chi-Square 0.3114

Z-Test of Wilson & Hilferty 0.4922

Hoelter Critical N 268

Root Mean Square Residual (RMSR) 0.1759

Standardized RMSR (SRMSR) 0.0586

Goodness of Fit Index (GFI) 0.9481

Parsimony Index Adjusted GFI (AGFI) 0.9292

Parsimonious GFI 0.7946

RMSEA Estimate 0.0173

RMSEA Lower 90% Confidence Limit

0.0000

RMSEA Upper 90% Confidence Limit 0.0410

Probability of Close Fit 0.9937

ECVI Estimate 0.7174

Page 103: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.7204

ECVI Upper 90% Confidence Limit 0.8406

Akaike Information Criterion 157.9945

Bozdogan CAIC 299.7336

Schwarz Bayesian Criterion 267.7336

McDonald Centrality 0.9869

Incremental Index Bentler Comparative Fit Index 0.0000

Bentler-Bonett NFI 0.1376

Bentler-Bonett Non-normed Index -0.7936

Bollen Normed Index Rho1 -0.0290

Bollen Non-normed Index Delta2 0.7144

James et al. Parsimonious NFI 0.1153

Page 104: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter

Estimate

attitude ---> X1 lam1 1.76418

attitude ---> X2 lam2 -0.18085

attitude ---> X3 lam3 -0.0005021

subnorms

---> X4 lam1 1.76418

subnorms

---> X5 lam2 -0.18085

subnorms

---> X6 lam3 -0.0005021

intention ---> X7 lam1 1.76418

intention ---> X8 lam2 -0.18085

intention ---> X9 lam3 -0.0005021

pastbeh ---> X10 lam1 1.76418

pastbeh ---> X11 lam2 -0.18085

pastbeh ---> X12 lam3 -0.0005021

beh ---> X13 lam1 1.76418

beh ---> X14 lam2 -0.18085

beh ---> X15 lam3 -0.0005021

attitude ---> intention _Parm1 4.38971

attitude ---> beh _Parm2 -0.02373

pastbeh ---> intention _Parm3 0.63146

pastbeh ---> beh _Parm4 0.03438

subnorms

---> intention _Parm5 14.59712

intention ---> beh _Parm6 0.00222

Page 105: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Variance Parameters

VarianceType

Variable Parameter Estimate

Exogenous attitude _Add01 0.48464

subnorms

_Add02 0.05162

pastbeh _Add03 0.83022

Error X1 _Add04 1.30799

X2 _Add05 2.80716

X3 _Add06 3.04478

X4 _Add07 2.87768

X5 _Add08 3.23934

X6 _Add09 2.84599

X7 _Add10 1.72009

X8 _Add11 3.05485

X9 _Add12 2.79922

X10 _Add13 0.20041

X11 _Add14 2.97714

X12 _Add15 2.87918

X13 _Add16 5.37074

X14 _Add17 3.30704

X15 _Add18 3.00446

intention _Add19 -1.06529

beh _Add20 -0.79000

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate

subnorms

attitude _Add21 -0.14446

pastbeh attitude _Add22 -0.07489

pastbeh subnorms _Add23 -0.01121

Page 106: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 1.30799 2.81637 0.5356

X10 0.20041 2.78434 0.9280

X11 2.97714 3.00429 0.00904

X12 2.87918 2.87918 7.27E-8

X13 5.37074 2.91631 -0.8416

X14 3.30704 3.28125 -0.00786

X15 3.00446 3.00446 -6.62E-8

X2 2.80716 2.82301 0.00562

X3 3.04478 3.04478 4.013E-8

X4 2.87768 3.03834 0.0529

X5 3.23934 3.24103 0.000521

X6 2.84599 2.84599 0

X7 1.72009 3.17673 0.4585

X8 3.05485 3.07015 0.00499

X9 2.79922 2.79922 4.216E-8

beh -0.79000 -0.78861 .

intention -1.06529 0.46802 .

Page 107: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter

Estimate

attitude ---> X1 lam1 0.73183

attitude ---> X2 lam2 -0.07493

attitude ---> X3 lam3 -0.0002003

subnorms

---> X4 lam1 0.22995

subnorms

---> X5 lam2 -0.02282

subnorms

---> X6 lam3 -0.0000676

intention ---> X7 lam1 0.67715

intention ---> X8 lam2 -0.07061

intention ---> X9 lam3 -0.0002053

pastbeh ---> X10 lam1 0.96334

pastbeh ---> X11 lam2 -0.09507

pastbeh ---> X12 lam3 -0.0002696

beh ---> X13 lam1 .

beh ---> X14 lam2 .

beh ---> X15 lam3 .

attitude ---> intention _Parm1 4.46699

attitude ---> beh _Parm2 .

pastbeh ---> intention _Parm3 0.84103

pastbeh ---> beh _Parm4 .

subnorms

---> intention _Parm5 4.84771

intention ---> beh _Parm6 .

Page 108: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate

Exogenous attitude _Add01 1.00000

subnorms _Add02 1.00000

pastbeh _Add03 1.00000

Error X1 _Add04 0.46442

X2 _Add05 0.99438

X3 _Add06 1.00000

X4 _Add07 0.94712

X5 _Add08 0.99948

X6 _Add09 1.00000

X7 _Add10 0.54147

X8 _Add11 0.99501

X9 _Add12 1.00000

X10 _Add13 0.07198

X11 _Add14 0.99096

X12 _Add15 1.00000

X13 _Add16 1.84162

X14 _Add17 1.00786

X15 _Add18 1.00000

intention _Add19 -2.27617

beh _Add20 .

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate

subnorms attitude _Add21 -0.91335

pastbeh attitude _Add22 -0.11807

pastbeh subnorms _Add23 -0.05416

Page 109: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simulation #8

Test: Model 2 (Intention Partially Mediates Attitude Effects on Behavior)

Data Set: Random

SAS SYNTAX:

proc calis data=SASUSER.SEM_RANDOMDATA; path attitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention beh, subnorms ---> intention, intention ---> beh;run;

SAS LOG:

1 proc calis data=SASUSER.SEM_RANDOMDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 beh ---> X13 X14 X15 = lam1 lam2 lam3,7 attitude ---> intention beh,8 subnorms ---> intention,9 intention ---> beh;10 run;

NOTE: Convergence criterion (GCONV=1E-8) satisfied.NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.WARNING: The estimated error variance for variable beh is negative.WARNING: The predicted variance of latent variable beh is negative.WARNING: The predicted variance of latent variable subnorms is negative.WARNING: Number of zero or negative variances encountered when computing standardized results: 2. Standardization with these values was not done.NOTE: PROCEDURE CALIS used (Total process time): real time 1.29 seconds cpu time 0.82 seconds

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_RANDOMDATA

N Records Read

228

Page 110: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Modeling Information

N Records Used 228

N Obs 228

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Latent beh intention

Exogenous Manifest

Latent attitude subnorms

Number of Endogenous Variables = 14Number of Exogenous Variables = 2

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 3.96942 1.67800

X2 X2 3.92882 1.68062

X3 X3 4.01325 1.74491

X4 X4 3.90928 1.74093

X5 X5 3.92321 1.80255

X6 X6 3.96071 1.68701

X7 X7 4.10400 1.78364

X8 X8 4.11832 1.75092

X9 X9 3.95523 1.67310

X13 X13

4.02703 1.70790

X14 X1 4.16736 1.81132

Page 111: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Simple Statistics

Variable Mean Std Dev

4

X15 X15

4.04412 1.73333

Page 112: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 -0.0003808

2 lam2 0.30562 0.00739

3 lam3 0.21152 0.01181

4 _Parm1 0.00100 0.00146

5 _Parm2 0.01000 0.00189

6 _Parm3 0.00100 -0.00824

7 _Parm4 0.01000 0.00138

8 _Add01 0.10000 0.00628

9 _Add02 0.27996 -0.0005856

10 _Add03 2.71570 -0.0001943

11 _Add04 2.81514 -0.0005373

12 _Add05 3.04025 0.0002559

13 _Add06 2.75089 0.0005664

14 _Add07 3.22303 -0.0002895

15 _Add08 2.83349 -0.0001012

16 _Add09 3.08138 -0.0005347

17 _Add10 3.05637 -0.0003439

18 _Add11 2.79478 -0.0002768

19 _Add12 2.74354 0.0003455

Page 113: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add13 3.26469 0.0003778

21 _Add14 2.99668 -0.0007404

22 _Add15 0.10000 0.01715

23 _Add16 0.17342 0.0000520

24 _Add17 0 0.07465

Value of Objective Function = 0.277001818

Page 114: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 24

Functions (Observations) 78

Optimization Start

Active Constraints 0 Objective Function

0.277001818

Max Abs Gradient Element 0.0746523857 Radius 1

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 0 5 0 0.25967 0.0173 0.0579 0.883 1.283

2 * 0 7 0 0.25226 0.00741 0.0467 0.194 0.435

3 * 0 9 0 0.24461 0.00765 0.00973 0.116 0.809

4 * 0 11 0 0.24129 0.00332 0.0100 0.0134 0.995

5 * 0 13 0 0.23544 0.00585 0.0238 0.00515

1.359

6 * 0 16 0 0.22401 0.0114 0.0207 0.00723

1.567

7 * 0 18 0 0.22238 0.00163 0.0809 0.00069

0.270

8 * 0 21 0 0.21575 0.00662 0.0137 0.0033 0.834

Page 115: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

7

9 * 0 24 0 0.21396 0.00180 0.0131 0.00245

0.865

10 * 0 26 0 0.21368 0.000272

0.0154 111E-16

0.460

11 * 0 28 0 0.21329 0.000391

0.00431 111E-16

0.891

12 * 0 30 0 0.21326 0.000029

0.00219 111E-16

0.449

13 * 0 32 0 0.21325 9.501E-6

0.00248 111E-16

0.258

14 * 0 34 0 0.21325 4.747E-6

0.00182 111E-16

0.190

15 * 0 36 0 0.21325 2.882E-6

0.00166 111E-16

0.166

16 * 0 38 0 0.21324 1.968E-6

0.00134 111E-16

0.159

17 * 0 40 0 0.21324 1.368E-6

0.00116 111E-16

0.155

18 * 0 42 0 0.21324 9.799E-7

0.000970

111E-16

0.155

19 * 0 44 0 0.21324 6.949E-7

0.000827

111E-16

0.154

20 * 0 46 0 0.21324 4.996E-7

0.000697

111E-16

0.154

21 * 0 48 0 0.21324 3.557E-7

0.000591

111E-16

0.154

22 * 0 50 0 0.21324 2.556E-7

0.000499

111E-16

0.154

23 * 0 52 0 0.21324 1.823E- 0.00042 111E- 0.154

Page 116: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

7 3 16

24 * 0 54 0 0.21324 1.308E-7

0.000358

111E-16

0.154

25 * 0 56 0 0.21324 9.342E-8

0.000303

111E-16

0.154

26 * 0 58 0 0.21324 6.702E-8

0.000256

111E-16

0.154

27 * 0 60 0 0.21324 4.789E-8

0.000217

111E-16

0.154

28 * 0 62 0 0.21324 3.433E-8

0.000183

111E-16

0.154

29 * 0 64 0 0.21324 2.454E-8

0.000155

111E-16

0.154

30 * 0 66 0 0.21324 1.759E-8

0.000131

111E-16

0.154

31 * 0 68 0 0.21324 1.258E-8

0.000111

111E-16

0.154

32 * 0 70 0 0.21324 9.012E-9

0.000094

111E-16

0.154

33 * 0 72 0 0.21324 6.447E-9

0.000079

111E-16

0.154

34 * 0 74 0 0.21324 4.619E-9

0.000067

111E-16

0.154

35 * 0 76 0 0.21324 3.304E-9

0.000057

111E-16

0.154

36 * 0 78 0 0.21324 2.366E-9

0.000048

111E-16

0.154

37 * 0 80 0 0.21324 1.693E-9

0.000041

111E-16

0.154

38 * 0 82 0 0.21324 1.212E- 0.00003 111E- 0.154

Page 117: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

9 4 16

39 * 0 84 0 0.21324 8.68E-10

0.000029

111E-16

0.154

40 * 0 86 0 0.21324 6.21E-10

0.000025

111E-16

0.154

41 * 0 88 0 0.21324 4.45E-10

0.000021

111E-16

0.154

42 * 0 90 0 0.21324 3.18E-10

0.000018

111E-16

0.154

43 * 0 92 0 0.21324 2.28E-10

0.000015

111E-16

0.154

44 * 0 94 0 0.21324 1.63E-10

0.000013

111E-16

0.154

Optimization Results

Iterations 44 Function Calls 97

Jacobian Calls 46 Active Constraints 0

Objective Function 0.2132392807 Max Abs Gradient Element 0.0000126347

Lambda 1.110223E-14 Actual Over Pred Change 0.1539224805

Radius 1.4102135198

Convergence criterion (GCONV=1E-8) satisfied.

Note: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.

Warning: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.

Page 118: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Note: Covariance matrix for the estimates is not full rank.

Note: The variance of some parameter estimates is zero or some parameter estimates are linearly related to other parameter estimates as shown in the following equations:

_Add02

= 1.368962

- 2.945907

* lam1

+ 0.909780

* lam2

+ 0.540043

* lam3

+ 0.620568

* _Add01

+ 1.677193

* _Add15

- 2.224321

* _Add16

- 1.469620

* _Add17

Page 119: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 12

N Moments 78

N Parameters 24

N Active Constraints 0

Baseline Model Function Value 0.2767

Baseline Model Chi-Square 62.8118

Baseline Model Chi-Square DF 66

Pr > Baseline Model Chi-Square 0.5885

Absolute Index Fit Function 0.2132

Chi-Square 48.4053

Chi-Square DF 54

Pr > Chi-Square 0.6891

Z-Test of Wilson & Hilferty -0.4939

Hoelter Critical N 339

Root Mean Square Residual (RMSR) 0.1598

Standardized RMSR (SRMSR) 0.0527

Goodness of Fit Index (GFI) 0.9656

Parsimony Index Adjusted GFI (AGFI) 0.9503

Parsimonious GFI 0.7900

RMSEA Estimate 0.0000

RMSEA Lower 90% Confidence Limit

0.0000

RMSEA Upper 90% Confidence Limit 0.0339

Probability of Close Fit 0.9971

ECVI Estimate 0.4375

Page 120: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.4766

ECVI Upper 90% Confidence Limit 0.5251

Akaike Information Criterion 96.4053

Bozdogan CAIC 202.7096

Schwarz Bayesian Criterion 178.7096

McDonald Centrality 1.0123

Incremental Index Bentler Comparative Fit Index .

Bentler-Bonett NFI 0.2294

Bentler-Bonett Non-normed Index -1.1448

Bollen Normed Index Rho1 0.0581

Bollen Non-normed Index Delta2 1.6349

James et al. Parsimonious NFI 0.1877

Page 121: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter Estimate StandardError

t Value

attitude ---> X1 lam1 1.21122 0.33128 3.65620

attitude ---> X2 lam2 -0.37406 0.19371 -1.93104

attitude ---> X3 lam3 -0.22204 0.16334 -1.35934

subnorms

---> X4 lam1 1.21122 0.33128 3.65620

subnorms

---> X5 lam2 -0.37406 0.19371 -1.93104

subnorms

---> X6 lam3 -0.22204 0.16334 -1.35934

intention ---> X7 lam1 1.21122 0.33128 3.65620

intention ---> X8 lam2 -0.37406 0.19371 -1.93104

intention ---> X9 lam3 -0.22204 0.16334 -1.35934

beh ---> X13 lam1 1.21122 0.33128 3.65620

beh ---> X14 lam2 -0.37406 0.19371 -1.93104

beh ---> X15 lam3 -0.22204 0.16334 -1.35934

attitude ---> intention _Parm1 -0.66977 0.90285 -0.74183

attitude ---> beh _Parm2 0.14388 0.48911 0.29418

subnorms

---> intention _Parm3 -0.07322 0.46707 -0.15677

intention ---> beh _Parm4 0.05295 0.38381 0.13797

Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 0.12757 0.31536 0.40453

subnorms _Add02 -0.54084 0.34353 -1.57436

Page 122: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Error X1 _Add03 2.60892 0.53544 4.87246

X2 _Add04 2.80723 0.26777 10.48373

X3 _Add05 3.06080 0.28804 10.62645

X4 _Add06 3.86665 0.90626 4.26657

X5 _Add07 3.32040 0.32280 10.28615

X6 _Add08 2.84533 0.26991 10.54195

X7 _Add09 2.62145 0.70070 3.74116

X8 _Add10 3.03617 0.29324 10.35405

X9 _Add11 2.77969 0.26322 10.56042

X13 _Add12 3.59598 0.79134 4.54418

X14 _Add13 3.32396 0.31931 10.40993

X15 _Add14 3.03531 0.28589 10.61701

intention _Add15 0.34479 0.40812 0.84484

beh _Add16 -0.45727 0.33890 -1.34929

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude

_Add17 -0.30212 0.16842 -1.79384

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 2.60892 2.79608 0.0669

X13 3.59598 2.92912 -0.2277

X14 3.32396 3.26036 -0.0195

X15 3.03531 3.01290 -0.00744

X2 2.80723 2.82508 0.00632

X3 3.06080 3.06709 0.00205

Page 123: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X4 3.86665 3.07320 -0.2582

X5 3.32040 3.24473 -0.0233

X6 2.84533 2.81867 -0.00946

X7 2.62145 3.16351 0.1713

X8 3.03617 3.08787 0.0167

X9 2.77969 2.79791 0.00651

beh -0.45727 -0.45456 .

intention 0.34479 0.36949 0.0668

Page 124: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter Estimate StandardError

t Value

attitude ---> X1 lam1 0.25872 0.33272 0.77760

attitude ---> X2 lam2 -0.07949 0.10900 -0.72926

attitude ---> X3 lam3 -0.04528 0.06675 -0.67839

subnorms

---> X4 lam1 .

subnorms

---> X5 lam2 .

subnorms

---> X6 lam3 .

intention ---> X7 lam1 0.41394 0.25209 1.64205

intention ---> X8 lam2 -0.12939 0.08729 -1.48232

intention ---> X9 lam3 -0.08069 0.06929 -1.16460

beh ---> X13 lam1 .

beh ---> X14 lam2 .

beh ---> X15 lam3 .

attitude ---> intention _Parm1 -0.39355 0.37836 -1.04015

attitude ---> beh _Parm2 .

subnorms

---> intention _Parm3 .

intention ---> beh _Parm4 .

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate

StandardError

t Value

Exogenous

attitude _Add01 1.00000

Page 125: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate

StandardError

t Value

subnorms _Add02 1.00000

Error X1 _Add03 0.93306 0.17216 5.41970

X2 _Add04 0.99368 0.01733 57.34352

X3 _Add05 0.99795 0.00605 165.06313

X4 _Add06 1.25818 0.26537 4.74130

X5 _Add07 1.02332 0.02376 43.06236

X6 _Add08 1.00946 0.01366 73.88386

X7 _Add09 0.82865 0.20870 3.97054

X8 _Add10 0.98326 0.02259 43.52659

X9 _Add11 0.99349 0.01118 88.85385

X13 _Add12 1.22767 0.24455 5.02002

X14 _Add13 1.01951 0.02254 45.22734

X15 _Add14 1.00744 0.01161 86.80159

intention _Add15 0.93316 0.58707 1.58952

beh _Add16 .

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add17 .

Page 126: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

SIMULATION #9

Tested: Model 1 (Theory of Reasoned Action)

Data Set: Random

SAS SYNTAX:

proc calis data=SASUSER.SEM_RANDOMDATA;pathattitude ---> X1 X2 X3 = lam1 lam2 lam3, subnorms ---> X4 X5 X6 = lam1 lam2 lam3, intention ---> X7 X8 X9 = lam1 lam2 lam3, beh ---> X13 X14 X15 = lam1 lam2 lam3, attitude ---> intention, subnorms ---> intention, intention ---> beh;run;

SAS LOG:

1 proc calis data=SASUSER.SEM_RANDOMDATA;2 path3 attitude ---> X1 X2 X3 = lam1 lam2 lam3,4 subnorms ---> X4 X5 X6 = lam1 lam2 lam3,5 intention ---> X7 X8 X9 = lam1 lam2 lam3,6 beh ---> X13 X14 X15 = lam1 lam2 lam3,7 attitude ---> intention,8 subnorms ---> intention,9 intention ---> beh;10 run;

NOTE: Convergence criterion (GCONV=1E-8) satisfied.NOTE: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.WARNING: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.WARNING: The estimated error variance for variable beh is negative.WARNING: The predicted variance of latent variable beh is negative.WARNING: The predicted variance of latent variable subnorms is negative.WARNING: Number of zero or negative variances encountered when computing standardized results: 2. Standardization with these values was not done.NOTE: PROCEDURE CALIS used (Total process time): real time 1.19 seconds cpu time 0.79 seconds

SAS OUTPUT:

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Model and Initial Values

Modeling Information

Data Set SASUSER.SEM_RANDOMDATA

N Records Read

228

Page 127: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Modeling Information

N Records Used 228

N Obs 228

Model Type PATH

Analysis Covariances

Variables in the Model

Endogenous

Manifest X1 X13 X14 X15 X2 X3 X4 X5 X6 X7 X8 X9

Latent beh intention

Exogenous Manifest

Latent attitude subnorms

Number of Endogenous Variables = 14Number of Exogenous Variables = 2

Page 128: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Descriptive Statistics

Simple Statistics

Variable Mean Std Dev

X1 X1 3.96942 1.67800

X2 X2 3.92882 1.68062

X3 X3 4.01325 1.74491

X4 X4 3.90928 1.74093

X5 X5 3.92321 1.80255

X6 X6 3.96071 1.68701

X7 X7 4.10400 1.78364

X8 X8 4.11832 1.75092

X9 X9 3.95523 1.67310

X13 X13

4.02703 1.70790

X14 X14

4.16736 1.81132

X15 X15

4.04412 1.73333

Page 129: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Initial Estimation Methods

1 Instrumental Variables Method

2 McDonald Method

3 Two-Stage Least Squares

Optimization StartParameter Estimates

N Parameter Estimate Gradient

1 lam1 1.00000 -0.0004157

2 lam2 0.30562 0.00740

3 lam3 0.21152 0.01179

4 _Parm1 0.00100 0.00144

5 _Parm2 0.00100 -0.00824

6 _Parm3 0.01000 0.00138

7 _Add01 0.10000 0.00610

8 _Add02 0.27996 -0.0005856

9 _Add03 2.71570 -0.0001919

10 _Add04 2.81514 -0.0005327

11 _Add05 3.04025 0.0002554

12 _Add06 2.75089 0.0005664

13 _Add07 3.22303 -0.0002895

14 _Add08 2.83349 -0.0001012

15 _Add09 3.08138 -0.0005347

16 _Add10 3.05637 -0.0003439

17 _Add11 2.79478 -0.0002768

18 _Add12 2.74354 0.0003539

19 _Add13 3.26469 0.0003723

Page 130: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Optimization StartParameter Estimates

N Parameter Estimate Gradient

20 _Add14 2.99668 -0.0007385

21 _Add15 0.10000 0.01715

22 _Add16 0.17341 0.0000562

23 _Add17 0 0.07448

Value of Objective Function = 0.2769830854

Page 131: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates 23

Functions (Observations) 78

Optimization Start

Active Constraints 0 Objective Function 0.2769830854

Max Abs Gradient Element 0.0744788597 Radius 1

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

1 * 0 5 0 0.25876 0.0182 0.0657 0.855 1.384

2 * 0 7 0 0.24714 0.0116 0.0300 0.141 1.209

3 * 0 9 0 0.24144 0.00570 0.0224 0.0207 0.710

4 * 0 11 0 0.23799 0.00345 0.0207 0.0222 0.571

5 * 0 13 0 0.23458 0.00341 0.0220 0.0127 0.790

6 * 0 16 0 0.23256 0.00202 0.0200 0.0188 0.570

7 * 0 18 0 0.23087 0.00169 0.0299 0.00799

0.772

8 * 0 20 0 0.22858 0.00229 0.0428 0.00345

0.661

9 * 0 22 0 0.22269 0.00589 0.0415 0.00123

1.271

Page 132: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

10 * 0 25 0 0.21811 0.00458 0.0240 0.00766

1.193

11 * 0 27 0 0.21696 0.00114 0.0747 111E-16

0.503

12 * 0 29 0 0.21437 0.00260 0.0173 0.00015

0.991

13 * 0 31 0 0.21382 0.000544

0.0104 111E-16

0.917

14 * 0 33 0 0.21364 0.000181

0.00210 111E-16

1.016

15 * 0 35 0 0.21362 0.000017

0.000593

111E-16

0.924

16 * 0 37 0 0.21362 3.702E-6

0.000748

111E-16

0.732

17 * 0 39 0 0.21362 9.365E-7

0.000378

111E-16

0.502

18 * 0 41 0 0.21362 3.058E-7

0.000381

111E-16

0.335

19 * 0 43 0 0.21362 1.295E-7

0.000263

111E-16

0.248

20 * 0 45 0 0.21362 6.904E-8

0.000227

111E-16

0.214

21 * 0 47 0 0.21362 4.135E-8

0.000173

111E-16

0.201

22 * 0 49 0 0.21362 2.605E-8

0.000142

111E-16

0.197

23 * 0 51 0 0.21362 1.674E-8

0.000113

111E-16

0.195

24 * 0 53 0 0.21362 1.082E-8

0.000092

111E-16

0.195

Page 133: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Iteration

Restarts

Function

Calls

ActiveConstrain

ts

Objective

Function

Objective

Function

Change

Max Abs

Gradient

Element

Lambda

RatioBetween

Actualand

Predicted

Change

25 * 0 55 0 0.21362 7.013E-9

0.000073

111E-16

0.195

26 * 0 57 0 0.21362 4.548E-9

0.000060

111E-16

0.194

27 * 0 59 0 0.21362 2.952E-9

0.000048

111E-16

0.194

28 * 0 61 0 0.21362 1.915E-9

0.000039

111E-16

0.194

29 * 0 63 0 0.21362 1.243E-9

0.000031

111E-16

0.194

30 * 0 65 0 0.21362 8.07E-10

0.000025

111E-16

0.194

31 * 0 67 0 0.21362 5.23E-10

0.000020

111E-16

0.194

32 * 0 69 0 0.21362 3.4E-10 0.000016

111E-16

0.194

33 * 0 71 0 0.21362 2.2E-10 0.000013

111E-16

0.194

34 * 0 73 0 0.21362 1.43E-10

0.000011

111E-16

0.194

Optimization Results

Iterations 34 Function Calls 76

Jacobian Calls 36 Active Constraints 0

Objective Function 0.2136193968 Max Abs Gradient Element 0.000010588

Lambda 1.110223E-14 Actual Over Pred Change 0.1944068834

Radius 70.12934888

Convergence criterion (GCONV=1E-8) satisfied.

Page 134: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Note: The Moore-Penrose inverse is used in computing the covariance matrix for parameter estimates.

Warning: Standard errors and t values might not be accurate with the use of the Moore-Penrose inverse.

Note: Covariance matrix for the estimates is not full rank.

Note: The variance of some parameter estimates is zero or some parameter estimates are linearly related to other parameter estimates as shown in the following equations:

_Add02

= 0.979535

- 3.012506

* lam1

+ 0.907608

* lam2

+ 0.561023

* lam3

+ 0.597522

* _Add01

+ 1.836001

* _Add15

- 2.445607

* _Add16

- 1.641415

* _Add17

Page 135: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Fit Summary

Modeling Info N Observations 228

N Variables 12

N Moments 78

N Parameters 23

N Active Constraints 0

Baseline Model Function Value 0.2767

Baseline Model Chi-Square 62.8118

Baseline Model Chi-Square DF 66

Pr > Baseline Model Chi-Square 0.5885

Absolute Index Fit Function 0.2136

Chi-Square 48.4916

Chi-Square DF 55

Pr > Chi-Square 0.7199

Z-Test of Wilson & Hilferty -0.5832

Hoelter Critical N 344

Root Mean Square Residual (RMSR) 0.1600

Standardized RMSR (SRMSR) 0.0528

Goodness of Fit Index (GFI) 0.9656

Parsimony Index Adjusted GFI (AGFI) 0.9511

Parsimonious GFI 0.8046

RMSEA Estimate 0.0000

RMSEA Lower 90% Confidence Limit

0.0000

RMSEA Upper 90% Confidence Limit 0.0323

Probability of Close Fit 0.9979

ECVI Estimate 0.4286

Page 136: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Fit Summary

ECVI Lower 90% Confidence Limit 0.4720

ECVI Upper 90% Confidence Limit 0.5155

Akaike Information Criterion 94.4916

Bozdogan CAIC 196.3666

Schwarz Bayesian Criterion 173.3666

McDonald Centrality 1.0144

Incremental Index Bentler Comparative Fit Index .

Bentler-Bonett NFI 0.2280

Bentler-Bonett Non-normed Index -1.4497

Bollen Normed Index Rho1 0.0736

Bollen Non-normed Index Delta2 1.8332

James et al. Parsimonious NFI 0.1900

Page 137: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

PATH List

Path Parameter Estimate StandardError

t Value

attitude ---> X1 lam1 1.17744 0.34358 3.42696

attitude ---> X2 lam2 -0.35474 0.18472 -1.92041

attitude ---> X3 lam3 -0.21928 0.15782 -1.38937

subnorms

---> X4 lam1 1.17744 0.34358 3.42696

subnorms

---> X5 lam2 -0.35474 0.18472 -1.92041

subnorms

---> X6 lam3 -0.21928 0.15782 -1.38937

intention ---> X7 lam1 1.17744 0.34358 3.42696

intention ---> X8 lam2 -0.35474 0.18472 -1.92041

intention ---> X9 lam3 -0.21928 0.15782 -1.38937

beh ---> X13 lam1 1.17744 0.34358 3.42696

beh ---> X14 lam2 -0.35474 0.18472 -1.92041

beh ---> X15 lam3 -0.21928 0.15782 -1.38937

attitude ---> intention _Parm1 -0.71581 1.08562 -0.65935

subnorms

---> intention _Parm2 -0.04826 0.55753 -0.08656

intention ---> beh _Parm3 -0.02212 0.28329 -0.07808

Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

Exogenous attitude _Add01 0.11677 0.36775 0.31753

subnorms _Add02 -0.58860 0.35766 -1.64568

Error X1 _Add03 2.63311 0.57845 4.55203

Page 138: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Variance Parameters

VarianceType

Variable Parameter Estimate StandardError

t Value

X2 _Add04 2.81145 0.26832 10.47787

X3 _Add05 3.06136 0.28813 10.62476

X4 _Add06 3.89180 0.92787 4.19433

X5 _Add07 3.31648 0.32214 10.29522

X6 _Add08 2.84726 0.27038 10.53049

X7 _Add09 2.61593 0.70941 3.68745

X8 _Add10 3.03813 0.29300 10.36891

X9 _Add11 2.77880 0.26325 10.55585

X13 _Add12 3.59033 0.79722 4.50355

X14 _Add13 3.32161 0.31849 10.42936

X15 _Add14 3.03578 0.28598 10.61536

intention _Add15 0.35880 0.44912 0.79890

beh _Add16 -0.47793 0.36058 -1.32546

Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude

_Add17 -0.32077 0.18793 -1.70692

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X1 2.63311 2.79500 0.0579

X13 3.59033 2.92800 -0.2262

X14 3.32161 3.26149 -0.0184

X15 3.03578 3.01281 -0.00762

X2 2.81145 2.82614 0.00520

X3 3.06136 3.06698 0.00183

X4 3.89180 3.07577 -0.2653

Page 139: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Squared Multiple Correlations

Variable Error Variance Total Variance R-Square

X5 3.31648 3.24241 -0.0228

X6 2.84726 2.81896 -0.0100

X7 2.61593 3.16369 0.1731

X8 3.03813 3.08785 0.0161

X9 2.77880 2.79779 0.00679

beh -0.47793 -0.47774 .

intention 0.35880 0.39510 0.0919

Page 140: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

The SAS System

The CALIS ProcedureCovariance Structure Analysis: Maximum Likelihood Estimation

Standardized Results for PATH List

Path Parameter Estimate StandardError

t Value

attitude ---> X1 lam1 0.24067 0.39178 0.61428

attitude ---> X2 lam2 -0.07211 0.12245 -0.58888

attitude ---> X3 lam3 -0.04279 0.07595 -0.56337

subnorms

---> X4 lam1 .

subnorms

---> X5 lam2 .

subnorms

---> X6 lam3 .

intention ---> X7 lam1 0.41610 0.25438 1.63574

intention ---> X8 lam2 -0.12689 0.08653 -1.46642

intention ---> X9 lam3 -0.08240 0.06996 -1.17786

beh ---> X13 lam1 .

beh ---> X14 lam2 .

beh ---> X15 lam3 .

attitude ---> intention _Parm1 -0.38914 0.38550 -1.00947

subnorms

---> intention _Parm2 .

intention ---> beh _Parm3 .

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate

StandardError

t Value

Exogenous

attitude _Add01 1.00000

subnorms _Add02 1.00000

Page 141: tamarastimatzecom.files.wordpress.com€¦  · Web viewSEM Simulations. SIMULATION #1. Tested: Model 3 (Intention Partial Mediates the Effects of Attitude & Past Behavior on Behavior)

Standardized Results for Variance Parameters

VarianceType

Variable Parameter Estimate

StandardError

t Value

Error X1 _Add03 0.94208 0.18858 4.99566

X2 _Add04 0.99480 0.01766 56.33427

X3 _Add05 0.99817 0.00650 153.58837

X4 _Add06 1.26531 0.27234 4.64614

X5 _Add07 1.02284 0.02344 43.64346

X6 _Add08 1.01004 0.01419 71.16430

X7 _Add09 0.82686 0.21169 3.90594

X8 _Add10 0.98390 0.02196 44.80317

X9 _Add11 0.99321 0.01153 86.14484

X13 _Add12 1.22620 0.24698 4.96480

X14 _Add13 1.01843 0.02176 46.79654

X15 _Add14 1.00762 0.01179 85.46239

intention _Add15 0.90813 0.67945 1.33657

beh _Add16 .

Standardized Results for Covariances Among Exogenous Variables

Var1 Var2 Parameter Estimate StandardError

t Value

subnorms attitude _Add17 .