forum.smartpls.com

hi!!

i have a categorical IV and have finished interpreting the data. the last thing I need to do is interpret q-squared. when I run the predictive model the q-squares are all negative. but when I run them unstandardised they are well above 0.

which version (standardised/unstandardised) should i interpret?

i also thought that due to the categorical IV - q-square and predictive relevance cannot be properly measured and thats why they are showing up negative. please let me know! really would appreciate any help

Yes, if you have a binary IV the predictive results are not correct for standardized results. We have fixed the issue and it will be correct with the next release. The unstandardized results are correct.

thank you for your response! really appreciate it