An analysis of the four food groups using multidimensional scaling

Abstract

To date, researchers have not tested whether the food groupings represented by various food guides reflect consumers' food classification systems. We undertook this study to explore whether multidimensional scaling (MDS) was a useful technique for examining respondents' perceived similarity among foods representing the Four Food Groups guide. This technique allows us to address the following questions: If individuals are given foods representative of the Four Food Groups, how would they group the foods? And, given the individuals' food groupings, what are the underlying dimensions or criteria on which they grouped the foods? Fifty-one college students rated all possible pairs of 23 foods, representing the Four Food Groups, on similar/dissimilar scales. Using an ALSCAL MDS analysis, we found that respondents grouped foods on 3 dimensions. Although the respondents used dimensions (or criteria) other than the nutrient composition of foods, they grouped the foods in a manner similar to but more complex than the Four Food Groups. The respondents seemed to perceive similarities among foods based on convenience in preparation, health-related properties, and source (animal and plant). Our results indicate that the MDS technique seems to be an appropriate method by which to examine whether the food groupings represented by a food guide reflects consumers' food classification systems.