Abstract
This paper compares the internal validity, external validity and ease of implementation of different types of compositional, decompositional and hybrid evaluation models in the context of recreational trip behavior. The results indicate that the full‐profile conjoint measurement model outperforms the self‐explicated models, the hybrid evaluation models and the trade‐off conjoint measurement model in terms of both internal and external validity, although the models differ less in terms of external validity. However, in terms of ease of implementation, the decompositional models may be less attractive.