SmartPLS 3.x - 3rd Order HCM - Methodology

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
pathaks
PLS Junior User
Posts: 1
Joined: Sun Jan 09, 2022 4:43 pm
Real name and title: pathaks

SmartPLS 3.x - 3rd Order HCM - Methodology

Post by pathaks »

Hi - Have a query on SmartPLS 3.x - 3rd Order HCM methodology based on the below model specification:

Independent : 1st Order HOC(Reflective) [#8] ||| 2nd Order HOC (Formative)[2#] ||| 3rd Order HOC - Formative [1#]
Dependent : All Reflective (First Order)

Methodology I am following (Need advice):

Step-1 (PLS):
1a- For the 1st Order (Reflective), measurement model validation as per Hair et al. (2014). Indicators repeated for 2nd and 3rd order HOC.

Step2 :
2a - Use the 8 LV scores from Step 1, with the all dependent variables (using original data) in the revised model specification.
2b - Added specific LV scores for each 2nd Order HOC and all 8 LV scores to 3rd Order HOC (repeat indicator) [===>>> IS THIS CORRECT ???). Do we need to repeat LV scores in this step???]
2c - Run SmartPLS Bootstrapping and measurement validation as per the guidelines by Hair et al (2014) for formative constructs.
2d - Get the LV scores (2#) again

Step3 :
3a - Repeat as Step 2 for 3rd Order HOC.

Step4 :
- Perform Structural Model Analysis

Unable to find a good reference for 3rd Order HCM (Reflective - Formative - Formative). The only reference i could find was the approach mentioned in James Gaskin Video here -> https://www.youtube.com/watch?v=LRND-H-hQQw in which the Step #2 and Step #3 are not done. Is that the correct approach or we need to perform Step 2 and Step 3 as mentioned above.

Any suggestions / guidance will be truly appreciated.

Regards.
Post Reply