STATISTICAL PACKAGE CLUSTERING MAY NOT BE BEST FOR GROUPING CONSUMERS TO UNDERSTAND THEIR MOST LIKED PRODUCTS: ISSUES WITH CLUSTER ANALYSIS

摘要: Ensuring that new products satisfy specific groups of consumers can impact successful product development. In sensory studies, cluster analysis has been used to segment consumers. Researchers often analyze mean values for consumer segments, presuming the segmented like or dislike similar products. This study investigates how well most/least liked match individual members clusters using various methods in two studies. Four statistical package clustering (SPC) were with hedonic data and transformed ranks. Next, most frequently rated/ranked highest each examined. manual extracted compared results SPC methods. Standard was not found separate appropriately understand their ranking/rating For these data, additional necessary produce segments where within group had same highest/lowest scoring products. PRACTICAL APPLICATIONS Statistical is a common method determining clusters, but it may be best separating understanding least The findings from this research show assumption will containing who have false. important because shows typical homogeneous researchers would obtain. further analyses optimization preference mapping. recommended more segments. addition, should developed generate are homogeneous.