A Proposal for Improving the Performance of Adaptive Conjoint Analysis

摘要: The identification of relevant attributes is the first objective conjoint analysis. Today, as a result technological development, it common for researchers to use adaptive analysis (ACA) which combines different types research (e.g. self-assessment questionnaires with an orthogonal design experiments). ACA, based on partial profiles, flexible sequential model that tailors experimental each respondent depending their previously stated preferences ordered in questionnaire. However, many authors hold full profile offers more advantages than one, because develops realistic description stimuli. Based this study proposes new strategy improve performance second step ACA process. This allows estimations main factors and two-factor interactions lowest number profiles. Our proposal approach profiles are arranged two-level factorial designs blocks two, levels factor codified vector manner.