Hi, a simple question:
I need cross loadings since I want to check discriminant validity.
How can I get them in PLS (or should I perform a Factor Analysis to get them)?
thanks a lot
sandra
cross loadings - discriminant validity
-
- PLS Expert User
- Posts: 139
- Joined: Wed Jul 05, 2006 1:43 pm
- Real name and title:
- Location: Coimbatore, India
cross loadings for discriminant validity
You get the cross loading chart as a part of smartPLS output tables. But to check discriminant validity we use another method. Find the individual AVE scores of the two constructs you are checking. Also note their construct correlation coeff. The RMS value of the AVE values should be larger than the construct correlation for good discriminant validity. Yo need not check individual indicator cross loadings.
-
- PLS Senior User
- Posts: 24
- Joined: Fri Nov 04, 2005 5:54 pm
- Real name and title: Dr. Joachim Schroer
- Contact:
Hello,
so far the only way I know for inspecting the cross-loadings is to use SPSS. IMHO, it might be more appropriate to use ordinal correlations (Kendall, Spearman) instead of Pearson if you want to stick to non-parametric tests.
Best wishes,
Joachim
so far the only way I know for inspecting the cross-loadings is to use SPSS. IMHO, it might be more appropriate to use ordinal correlations (Kendall, Spearman) instead of Pearson if you want to stick to non-parametric tests.
Best wishes,
Joachim
Dr. Joachim Schroer
PRIOTAS GmbH
Hohenzollernring 72
50672 Köln
http://www.priotas.de/
Feedback to progress
PRIOTAS GmbH
Hohenzollernring 72
50672 Köln
http://www.priotas.de/
Feedback to progress
-
- PLS Expert User
- Posts: 139
- Joined: Wed Jul 05, 2006 1:43 pm
- Real name and title:
- Location: Coimbatore, India
cross loadings
Yes, I repeat, you do get cross loadings in smartPLS outputs. The tables have the following order when you check the html outputs
1. Data pre processing
2. Model
3. Index values
4. PLS
a. Quality criteria
i. Latent variable correlations
ii. Overview
iii. Redundancy
iv. CROSS LOADINGS
Check it out. You are not blind! Just look properly
1. Data pre processing
2. Model
3. Index values
4. PLS
a. Quality criteria
i. Latent variable correlations
ii. Overview
iii. Redundancy
iv. CROSS LOADINGS
Check it out. You are not blind! Just look properly
-
- PLS Expert User
- Posts: 139
- Joined: Wed Jul 05, 2006 1:43 pm
- Real name and title:
- Location: Coimbatore, India
cross loadings
No, I didn't do anything special. I am attaching a sample from a two construct moel I ran:
Table of contents (complete)
* Data Preprocessing
o Results (chronologically)
+ Step 0 (Original Matrix)
* Model
o Specification
+ Manifest Variable Scores (Original)
+ Measurement Model Specification
+ Structural Model Specification
* Index Values
o Results
+ Path Coefficients
+ Measurement Model (restandardised)
+ Measurement Model
+ Latent Variable Scores (unstandardised)
+ Index Values for Latent Variables
* PLS
o Quality Criteria
+ Latent Variable Correlations
+ Overview
+ Redundancy
+ Cross Loadings
+ Cronbachs Alpha
+ R Square
+ Total Effects
+ AVE
+ Communality
+ Composite Reliability
o Calculation Results
+ Manifest Variable Scores (Used)
+ Path Coefficients
+ Outer Weights
+ Outer Loadings
+ Stop Criterion Changes
+ Latent Variable Scores
+ Outer Model (Weights or Loadings)
I hope this is useful
Table of contents (complete)
* Data Preprocessing
o Results (chronologically)
+ Step 0 (Original Matrix)
* Model
o Specification
+ Manifest Variable Scores (Original)
+ Measurement Model Specification
+ Structural Model Specification
* Index Values
o Results
+ Path Coefficients
+ Measurement Model (restandardised)
+ Measurement Model
+ Latent Variable Scores (unstandardised)
+ Index Values for Latent Variables
* PLS
o Quality Criteria
+ Latent Variable Correlations
+ Overview
+ Redundancy
+ Cross Loadings
+ Cronbachs Alpha
+ R Square
+ Total Effects
+ AVE
+ Communality
+ Composite Reliability
o Calculation Results
+ Manifest Variable Scores (Used)
+ Path Coefficients
+ Outer Weights
+ Outer Loadings
+ Stop Criterion Changes
+ Latent Variable Scores
+ Outer Model (Weights or Loadings)
I hope this is useful