Help using Base SAS procedures

Path analysis

Reply
Contributor mei
Contributor
Posts: 62

Path analysis

Dear Sir,

Expected relationships as follows:

Direct effect:  Idiorisk -----> advertising

Indirect effect: Idiorisk --> Gindex ---> Advertising

This is the program for path analysis:

proc calis data=testP1;

path  idiorisk ---> gindex,

idiorisk gindex --->  xad_log;

  run;

The report is shown after the q- it may not show good view but i have no choice as not allowed to attach word doc.

My q:

1.  See the final part of the report, where we can interpret the coefficient and t-statistics to see wht it is significant.

I can present the coefficient and t-stats for direct effect- that is from idiorisk to advertising-

idiorisk

--->

xad_log

_Parm2

-0.13823

  1. 0.01957

-7.06337

Now i have to present the mediated path ie the product of the indirect path.

idiorisk

--->

GINDEX

_Parm1

-0.05267

  1. 0.02013

-2.61637

GINDEX

--->

xad_log

_Parm3

0.15389    0.01952    7.88173

coefficient for the indirect path would be -0.05267 x 0.15389= 0.008

Q: how do i calculate the standard error and t-statistics for the joint-product of this indirect effect?

(Purpose: is to see wht the direct effect or indirect effect is more weighted?)

2. I read from SAS note from help menu that Chi-square should be insiginificant so that the model is not rejected. The standardized root mean squares of residuals (SRMSR) and the root mean square error of approximation (RMSEA) must be very small.

How do we know whether this model is a good model?

Thanks

Mei

Result:

The SAS System

The CALIS Procedure

Covariance Structure Analysis: Model and Initial Values

Data Set

  1. WORK.TESTP1

N Records Read

8256

N Records Used

2455

N Obs

2455

Model Type

PATH

Analysis

Covariances

Endogenous

Manifest

GINDEX xad_log

Latent

Exogenous

Manifest

idiorisk

Latent

Number of Endogenous Variables = 2
Number of Exogenous Variables = 1

idiorisk

--->

GINDEX

_Parm1

.

idiorisk

--->

xad_log

_Parm2

.

GINDEX

--->

xad_log

_Parm3

.

Exogenous

idiorisk

_Add1

.

Error

xad_log

_Add2

.

GINDEX

_Add3

.

NOTE: Parameters with prefix '_Add' are added by PROC CALIS.

 


 

The SAS System

The CALIS Procedure

Covariance Structure Analysis: Descriptive Statistics

idiorisk

  1. 1.51555
  2. 0.81104

xad_log

  1. 3.14352
  2. 2.18902

GINDEX

Governance Index (Gompers, Ishii, Metrick)

  1. 8.82566
  2. 2.46932

 


 

The SAS System

The CALIS Procedure

Covariance Structure Analysis: Optimization

1

Observed Moments of Variables

2

McDonald Method

3

Two-Stage Least Squares

1

_Parm1

-0.16036

-4.565E-18

2

_Parm2

-0.37309

  1. 8.2743E-19

3

_Parm3

  1. 0.13642
  2. 4.626E-17

4

_Add1

  1. 0.65778

-9.78E-36

5

_Add2

  1. 4.57604

-1.379E-18

6

_Add3

  1. 6.08063

-1.204E-19

Value of Objective Function = 0

 
 

 


 

The SAS System

The CALIS Procedure

Covariance Structure Analysis: Optimization

Levenberg-Marquardt Optimization

Scaling Update of More (1978)

Parameter Estimates

6

Functions (Observations)

6

Active Constraints

0

Objective Function

0

Max Abs Gradient Element

  1. 4.626042E-17

Radius

1

Iterations

0

Function Calls

4

Jacobian Calls

1

Active Constraints

0

Objective Function

0

Max Abs Gradient Element

  1. 4.626042E-17

Lambda

0

Actual Over Pred Change

0

Radius

1

Convergence criterion (ABSGCONV=0.00001) satisfied.

 


 

The SAS System

The CALIS Procedure

Covariance Structure Analysis: Maximum Likelihood Estimation

Modeling Info

N Observations

2455

N Variables

3

N Moments

6

N Parameters

6

N Active Constraints

0

Baseline Model Function Value

  1. 0.0489

Baseline Model Chi-Square

  1. 119.8827

Baseline Model Chi-Square DF

3

Pr > Baseline Model Chi-Square

<.0001

Absolute Index

Fit Function

  1. 0.0000

Chi-Square

  1. 0.0000

Chi-Square DF

0

Pr > Chi-Square

.

Z-Test of Wilson & Hilferty

.

Hoelter Critical N

.

Root Mean Square Residual (RMSR)

  1. 0.0000

Standardized RMSR (SRMSR)

  1. 0.0000

Goodness of Fit Index (GFI)

  1. 1.0000

Parsimony Index

Adjusted GFI (AGFI)

.

Parsimonious GFI

  1. 0.0000

RMSEA Estimate

.

Probability of Close Fit

.

ECVI Estimate

  1. 0.0049

ECVI Lower 90% Confidence Limit

.

ECVI Upper 90% Confidence Limit

.

Akaike Information Criterion

  1. 12.0000

Bozdogan CAIC

  1. 52.8353

Schwarz Bayesian Criterion

  1. 46.8353

McDonald Centrality

  1. 1.0000

Incremental Index

Bentler Comparative Fit Index

  1. 1.0000

Bentler-Bonett NFI

  1. 1.0000

Bentler-Bonett Non-normed Index

.

Bollen Normed Index Rho1

.

Bollen Non-normed Index Delta2

  1. 1.0000

James et al. Parsimonious NFI

  1. 0.0000

 
 

 


 

The SAS System

The CALIS Procedure

Covariance Structure Analysis: Maximum Likelihood Estimation

idiorisk

--->

GINDEX

_Parm1

-0.16036

  1. 0.06138

-2.61274

idiorisk

--->

xad_log

_Parm2

-0.37309

  1. 0.05332

-6.99743

GINDEX

--->

xad_log

_Parm3

  1. 0.13642
  2. 0.01751
  3. 7.78998

Exogenous

idiorisk

_Add1

  1. 0.65778
  2. 0.01878
  3. 35.02856

Error

xad_log

_Add2

  1. 4.57604
  2. 0.13064
  3. 35.02856

GINDEX

_Add3

  1. 6.08063
  2. 0.17359
  3. 35.02856

GINDEX

  1. 6.08063
  2. 6.09755
  3. 0.00277

xad_log

  1. 4.57604
  2. 4.79181
  3. 0.0450

The SAS System

The CALIS Procedure

Covariance Structure Analysis: Maximum Likelihood Estimation

idiorisk

--->

GINDEX

_Parm1

-0.05267

  1. 0.02013

-2.61637

idiorisk

--->

xad_log

_Parm2

-0.13823

  1. 0.01957

-7.06337

GINDEX

--->

xad_log

_Parm3

  1. 0.15389
  2. 0.01952
  3. 7.88173

Exogenous

idiorisk

_Add1

  1. 1.00000

Error

xad_log

_Add2

  1. 0.95497
  2. 0.00818
  3. 116.72457

GINDEX

_Add3

  1. 0.99723
  2. 0.00212
  3. 470.27502
Ask a Question
Discussion stats
  • 0 replies
  • 176 views
  • 0 likes
  • 1 in conversation