Modeling Yeah- and Nay-Saying to Alternatives in Conjoint Experiments

Using a series of hurdle choice models, this study considers both nay-saying and yeah-saying to alternatives offered in a conjoint experiment. These behaviors are characterized by respondents persistently choosing the no-choice alternative or choosing at least one of the non-empty options offered in a survey. Results show that jointly consider nay-saying and yeah-saying in a two-hurdle model drastically improves model fit; welfare implications based on hurdle models are also different from those based on models without hurdle specification.