Cannot calculate robust standard errors (vcovHC): multicollinearity and NaN error
After doing some reading and following @DominicComtois suggestion, I think here I will have to go with HC1. HC3, HC4 (Long and Ervin 2000, Hayes and Cai 2007), or HC4m (Cribari-Neto and Silva 2011) would be better, but they all give NaN results. The problem with HC0 is that it...