Endogeneity of indicator variables in hybrid choice models: Monte Carlo investigation vs. stated preference study