Higher Order Construct

This forum is closed, and read-only.
Locked
kfls_83
PLS Junior User
Posts: 6
Joined: Tue Dec 15, 2020 5:34 am
Real name and title: Khawaja Fawad Latif

Higher Order Construct

Post by kfls_83 »

I have a question pertinent to the Extended Repeated Indicators Approach in the paper titled "How to specify, estimate, and validate higher-order constructs in PLS-SEM".

Where in Embedded and Disjoint Two Stage Approach we take Latent Variable Score and use it to establish Discriminant Validity of Higher Order Construct and further use LVS to test our structural model, do we need to do this in Extended Repeated Indicators Approach.

If not, will this not affect the mediation or moderation results? Plus we will not do second stage to test for HOC in Extended Repeated Indicators Approach?

My other question is pertinent to Mode A and Mode B. If you can please refer to the model attached,

When using Embedded Approach for Reflect-Formative, in the second order (Repeated Indicators in this case OL and RS) (Its like X1 - X9 added to HOC) , shall i put my indicators as formative or reflective to draw LVS. As in the paper, it is recommended to use Mode B (Formative) but in a webinar by Prof. S. Mostafa, The Prof kept them reflective in the second order.

Last Question, When checking for Significance or outer weights to assess measurement model, shall we use path or factor in the bootstrapping option.
Attachments
Untitled.png
Untitled.png (159.24 KiB) Viewed 31523 times
Extended Repeated Indicators Approach.png
Extended Repeated Indicators Approach.png (37.46 KiB) Viewed 31523 times
Locked