摘要:The study combines different theoretical approaches in the field of conjoint analysis to estimate the im-portance of product related attributes. This is of major importance in food marketing, where we still try to find a valid answer, in particular, how to measure the real willingness to pay (WTP) for specific product specifica-tions. Based on a comprehensive literature analysis, a common method was used to approximate the im-portance of several product attributes. As usually suggested in literature, we used discrete choice modeling and developed a choice based experimental design considering selected product attributes. The study object was frozen pizza, a convenience good frequently bought by most households. Up to this point, there is nothing special about the choice based experiment in comparison to direct measure-ment of the importance of product attributes. However, one of the core problems of discrete choice modeling – the approximation of individual utility functions – was then addressed by transforming the choices of con-sumers into scores. With these scores traditional conjoint measurement can be used to approximate individual utilities even in choice based experiments. The individual part-worth utilities will be compared with a usual but very complex approach to approximate individual part-worth utilities, the hierarchical Bayes method. Our ap-proach addresses methodological considerations concerning the restrictions of discrete choice modeling, namely the complexity of approximating individual utilities which is of huge importance in particular for market segmentation.
其他摘要:The study combines different theoretical approaches in the field of conjoint analysis to estimate the im-portance of product related attributes. This is of major importance in food marketing, where we still try to find a valid answer, in particular, how to measure the real willingness to pay (WTP) for specific product specifica-tions. Based on a comprehensive literature analysis, a common method was used to approximate the im-portance of several product attributes. As usually suggested in literature, we used discrete choice modeling and developed a choice based experimental design considering selected product attributes. The study object was frozen pizza, a convenience good frequently bought by most households. Up to this point, there is nothing special about the choice based experiment in comparison to direct measure-ment of the importance of product attributes. However, one of the core problems of discrete choice modeling – the approximation of individual utility functions – was then addressed by transforming the choices of con-sumers into scores. With these scores traditional conjoint measurement can be used to approximate individual utilities even in choice based experiments. The individual part-worth utilities will be compared with a usual but very complex approach to approximate individual part-worth utilities, the hierarchical Bayes method. Our ap-proach addresses methodological considerations concerning the restrictions of discrete choice modeling, namely the complexity of approximating individual utilities which is of huge importance in particular for market segmentation.