Conventional conjoint analysis is a data analysis method based on experimental data collection, and is widely used in fields such as product and service development and marketing strategies. In conjoint analysis, experiments are often conducted using qualitative factors, which can lead to issues such as a lack of flexibility in policy design. In this study, we use virtual experiments designed to achieve the challenges of conventional conjoint analysis to collect experimental data that does not overlook combination effects (interactions) or quadratic terms.
In addition, while the main methods for evaluating experimental data include (A) the scoring method and (B) the ranking method, this study proposes (C) the difference apportionment method. Furthermore, to address the issue that collecting experimental data takes more time and effort than questionnaire surveys, we will attempt to collect experimental data completely online using videos, electronic files, and the Internet.
In this study, we discuss approaches to achieving traditional challenges and collecting high-quality experimental data based on 1) optimal planning using a predictive model table instead of an orthogonal design, 2) evaluation using a difference allocating method, and 3) virtual experiment examples incorporating online experiments.
