A SYSTEM FOR THE PREFERENCE EVALUATION OF CYCLIC MENUS

Abstract

A quantitative method to evaluate the overall preference of cyclic menus is presented. The model developed consists of two separate factors relating to menu preference rating. The first factor is the time-preference relationship for food items. From survey data, time-preference functions have been constructed which indicate how the hedonic rating of a food item varies with the time since the food was last served. The function assumes a quadratic form with entrees tending to be most time sensitive and salads and vegetables the least. The second factor is the meal component weights for the evaluaton of an entire meal based on the consumer hedonic preference values for the generic components of the meal (entree, starch, vegetable, salad, and dessert). Regression techniques have been applied to survey data to generate the set of importance factors, or meal component weights, for each component. For example, the results indicate that the entree represents about 49% of the total preference value of the meal. This approach has provided a basis for an additive linear model for overall meal preference. A general system for the overall evaluation of cyclic menus has also been developed based on application of an additive linear model. This procedure accounts for the maximum preference values for food items, the time-dependency of the food item preference values, and the meal component weighting factors and is easily adapted to the computer, as shown by a specific example. The overall system appears to offer a logical, realistic, and integrated approach which reflects time factors and the relative importances of the various components of meals in the evaluation of cyclic menus. Further research in this area is recommended to refine this model and test its validity in an operational environment.