I have nominal variables as both control variables as well as independent variables in my model. I have read the following 2 posts regarding this where it has been said that PLS can not accomodate nominal variables and later on it has also been said that you can dummy code and use it just like in regression.
Could you please let me know if this is possible and point me to any examples or journal articles which have done this. Can the interpretation of path coefficients be done the same way as in regression when you dummy code these variables.
If dependent variable used non-metric scale, acctually PLS cannot handle it but may be future is possible (current develop new algorithm).
Some paper try and effort for explain this issue, check Russolilo (2009 and 2012) Non-metric PLS in literature this forum.