forum.smartpls.com

Hi!

I hope someone can help me out and tell me how to deal with two independent variables (LV's) which I believe are correlated. Is there any way to take account of this correlation without drawing a path and thus implying a cause-effect relationship? I tried drawing a two paths (implying a recursive relationship...) but PLS does not accept that...

Thanks!

In LISREL (covariance based) you must draw the covariance, but in PLS you don´t need.

See viewtopic.php?t=125

IN the output you could see de matrix of correlations between LVs.

PLS minimize the error´s covariance (predictive efficient), LISREL minimize the diference between data and predicted matrix of covariance.

Best regards.