I am writing my PhD Thesis at the moment and made a SEM in SmartPLS including a 2nd Order Construct. In the Quality Criteria Section I can find almost for every variable the values. Thanks to this Forum I know, that I have to calculate AVE and Communality by hand for the 2nd Order Construct. But where lies the problem, if i get a blank/zero at r² for certain variables, including the 2nd Order construct?
I know you have 2nd construct type I or III, right?
If yes, R-square not show in second-order but first-order. Because direct link Type I or III from second-order to first-order.
Thx for the prompt reply. According to ringle et al. 2012 i have Type I reflective-reflective. So, which r Square do i have to take or what do i have to compute? Or Is it just not determined in this Case And thats ok?
Hi,
Yes, if position second-order construct Type I in endogenous, SmartPLS can compute R-square automatic (just for explain variance predictor, no construct dimension).
Different with Thom case above (position in exogenous).