I conducted a research to determine significant indicators of Hepatitis C among sanitary staff at hospitals. The indicators are related to demography, risk factors and organizational characteristics. I tried to do a binary logistic regression by coding all the indicators as categorical. However many of them have significant correlations among them. Thus I'm forced to use PLS. I have two questions:-

1. Can we use categorical variables in SmartPLS?

2. How should I determine the factors for PLS regression. Should I just assign the demographic indicators under one factor, risk factors under the second factor and the hospital xtics under the third? Or should I conducted Exploratory Factor Analysis? Remember that the latter might result in elimination of some of the indicators.