Thurstone's Pairwise Comparison Model to explore preferences of humans for different food items
L. Middeldorp (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Dorota Kurowicka – Mentor
G. F. Nane – Graduation committee member
JG Spandaw – Graduation committee member
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
The method of pairwise comparisons is a method that is commonly used in psychology to model human preferences for a set of objects. In a pairwise comparison experiment, the preferences of a sample are obtained by comparing the objects in pairs. Each individual in the sample judges every possible pair of objects and expresses which of the two object he prefers. Thurstone's Pairwise
Comparison Model can then be used on the gathered data in order to obtain rankings of the objects based on the preferences of the sample. This thesis investigates several properties of Thurstone's Pairwise Comparison Model and how it can be used to model human food preferences. In recent years, people have become more aware of the importance of eating healthy food. It may, however, be questioned whether healthy food is generally preferred by the population. This can be examined by means of Thurstone's Pairwise Comparison Model. In order to examine how the model performs, a data study has been carried out using two sets of food products. One set consisted of general food
products, while the other set was comprised of snack products. Both sets of objects consisted of both healthy and unhealthy food products. The influence of an introduction text, emphasizing the importance of healthy food, on the preferences of the individuals has been examined as well. For this, each set of food products was judged by two groups, each group receiving a different introduction text before starting the pairwise comparison experiment. The results of the data study yielded that Thurstone's Pairwise Comparison Model is more suitable for modelling preferences of humans for general food products. Furthermore, the introduction text turned out to be not of influence on the preferences of participants.