Which relations between LV are significant?

This forum is closed, and read-only.
Locked
bergsorensen
PLS User
Posts: 13
Joined: Wed Jun 25, 2008 10:40 am
Real name and title:

Which relations between LV are significant?

Post by bergsorensen »

maybe a dumb question - but new to PLS modelling - how can you tell if the relations between latent variables in a PLS model are significant? (besides from getting the estimates on the relations?)

Model explanations (r-square). I assume that each endogenous variable (for example in the ECSI model, loytlty satisfaction and perceived value, have an r-squre value defined/calculated while exgeneous do not?

Best regards

Kasper
Last edited by bergsorensen on Thu Jun 26, 2008 4:25 pm, edited 2 times in total.
bergsorensen
PLS User
Posts: 13
Joined: Wed Jun 25, 2008 10:40 am
Real name and title:

Post by bergsorensen »

I hope i understand this correct. The bootstrapping procedure is performed to test the significance of the relations. Telling if the relation is significant, is done by looking at the number computed by the bootstrapping and then looked up on a t-distribution.

If for example using 500 samples(infinite degrees of freedom) and cases = n (not actual observations), an t-number above 1,96 indicates an significant path? Is this correctly understod?

Best regards Kasper
Locked