levels by multiplying the importance weights and the respective attribute
level desirability scores. We augment the self-explicated data with data obtained from each respondent on a smaller set of complete product bundles.
Finally we compute the adjusted part-worth functions for each respondent.