Token-Based Conjoint evaluates large feature sets by asking respondents to select varying numbers of features into a bundle. This research tests whether standard additive TBC models overlook two sources of utility: baseline utility for the base product and bundle-size effects associated with k. A purpose-built study with fixed bundle-size holdouts will be used to compare additive, baseline, k-effects, and combined models. Attendees will gain practical guidance for modeling bundle size and evaluating marginal utility patterns in token-based conjoint studies.
