forum.smartpls.com

Hello everyone,

I have a question about running IPMA for my model.

My theoretical model is composed of 3 variables (cf. attached image) : one second-order construct (IV) with 3 first order (its dimensions) (reflective-reflective construct) and 2 other reflective constructs (DV1 and DV2). I intend to see which dimensions have high performance for predicting the target construct (DV2) by running IPMA. In Hair et al. (2018) Advanced issues in PLS-SEM, the instruction is mentionned for model of first-order construct where all predictors have a same scale.

With my model having 2nd order construct, should I use 2-stage approach to transform model then run IPMA for new model ? In this case, the construct IV is reflected by 3 manifest constructs (I use Latent variable scores for 3 dimensions), but these variables (dimensions) don't have the same range (min,max) because I use standardize score.

Could anyone advise me how to run IPMA for model having 2nd order construct ?

Thank you very much !

You have two options when you want to use the IPMA for a second-order construct.

1) You use the repeated-indicator approach. That should work just fine.

2) You could use the disjoint two-stage approach and take the necessary latent variable scores as unstandardized LV scores from an IPMA analysis on the first stage model.

Please also read:
https://www.researchgate.net/publicatio ... in_PLS-SEM

Following the discussion here, how can we find the IPMA results for the Low order constructs on IV?