Outer weight vs LV important index in IPMA

Frequently asked questions about PLS path modeling.
Post Reply
lilyv
PLS User
Posts: 11
Joined: Thu Apr 04, 2019 5:07 pm
Real name and title: Lily Vu

Outer weight vs LV important index in IPMA

Post by lilyv »

Hello everyone,
I have some questions about the assessment of my HOC model (higher-order model). My HOC model is reflective-formative with 6 dimensions.

- When I use repeated-indicator approach, the outer weights of dimensions I found in "Path coefficients results" is respectively : 0,30-0,25-0,20-0,19-0,22-0,17.
- When I use two-stage approach, the outer weights in step 2 is respectively : 0,25-0,19-0,12-0,19-0,21-0,37.
- When I run IPMA for HOC target, the LV important index for each dimension is respectively: 0,20-0,17-0,12-0,20-0,14-0,17.

As you see, each dimensions gets a different coefficient. My questions are follows :
1) Why there is a big different between outer weights and the LV important index in IPMA ? I think The LV important index (total effect) and outer weight are coefficient beta, is that right ?
2) If I want to keep report IPMA results, before that, I have to choose which approach to the validation of HOC model ? Is this conflict if I report the outer weights followed 2-stage approach then report IPMA results (using repeated indicator approach) in the same paper ?

Thanks in advance.
Post Reply