Randomized First Choice corrects for product similarity by simulating products multiple times, while each time adding error to the utility for each attribute and product concept. Usually attributes have correlated error terms applied, however some attributes would seem to not justify any correction for product similarity. Correcting for product similarity for the Price attribute is problematic, particularly leading to strange kinks in derived demand curves for pricing studies. We generally recommend you un-check the box for your price attribute. When an attribute is un-checked in this dialog, independent attribute error is drawn for this attribute across the product concepts within the simulation scenario.